{"comparison_contract":"Original and Clean are reconstructed from the same accepted attempt universe for every row.","release_id":"gpqa-clean-results-v0.1.0","rows":[{"clean":{"denominator":189,"numerator":184,"percent":"97.35%"},"clean_rank":1,"configuration_id":"gpqa_cfg_v1_7934de713f82e39e9592c19190f2c0c4b7a5afc288e8b146a975a0b4243d3076","delta":{"denominator":2079,"numerator":71,"percentage_points":"+3.42 pp"},"model_configuration":"gpt-5.5-pro-pre-release_xhigh","paired_original":{"denominator":33,"numerator":31,"percent":"93.94%"},"paired_original_rank":4,"protocol":"mean@7","rank_change":3},{"clean":{"denominator":189,"numerator":184,"percent":"97.35%"},"clean_rank":1,"configuration_id":"gpqa_cfg_v1_bfdb9cba805be9f524447f939b0d26c6c2a84b2eb93aa007c9f5c6d024b1fd8a","delta":{"denominator":2079,"numerator":71,"percentage_points":"+3.42 pp"},"model_configuration":"Gemini 3.6 Flash (high)","paired_original":{"denominator":33,"numerator":31,"percent":"93.94%"},"paired_original_rank":4,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":36,"numerator":35,"percent":"97.22%"},"clean_rank":3,"configuration_id":"gpqa_cfg_v1_524a33f2416077d9ee7c8a56c4e245d2737cfc16dc1505dcc72a7b1994afc88a","delta":{"denominator":528,"numerator":17,"percentage_points":"+3.22 pp"},"model_configuration":"gpt-5.5-pre-release_xhigh","paired_original":{"denominator":1584,"numerator":1489,"percent":"94.00%"},"paired_original_rank":3,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":168,"numerator":163,"percent":"97.02%"},"clean_rank":4,"configuration_id":"gpqa_cfg_v1_c0291f9a5fc3b09556c001225b7f3d6b3b1de8db4895ffa3e136dbc321bb3809","delta":{"denominator":1232,"numerator":31,"percentage_points":"+2.52 pp"},"model_configuration":"gemini-3.1-pro-preview","paired_original":{"denominator":528,"numerator":499,"percent":"94.51%"},"paired_original_rank":1,"protocol":"mean@8","rank_change":-3},{"clean":{"denominator":63,"numerator":61,"percent":"96.83%"},"clean_rank":5,"configuration_id":"gpqa_cfg_v1_e99e75b593bbdcaf766b1b732e2a1f92ac6b0acbdc8bd59190ce0972000f26e1","delta":{"denominator":1386,"numerator":47,"percentage_points":"+3.39 pp"},"model_configuration":"GPT-5.6 Sol (max)","paired_original":{"denominator":198,"numerator":185,"percent":"93.43%"},"paired_original_rank":6,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":756,"numerator":731,"percent":"96.69%"},"clean_rank":6,"configuration_id":"gpqa_cfg_v1_2f4d76e7d0e2db97f02626aaf55d382c339a5f505c3d382cb8492f9239d5894c","delta":{"denominator":33264,"numerator":1147,"percentage_points":"+3.45 pp"},"model_configuration":"gpt-5.4-2026-03-05_xhigh","paired_original":{"denominator":1584,"numerator":1477,"percent":"93.24%"},"paired_original_rank":7,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":27,"numerator":26,"percent":"96.30%"},"clean_rank":7,"configuration_id":"gpqa_cfg_v1_283043bba1e8117ffaa1f9aace7567853f92188831fd41efb8f9da8fb29f2f3f","delta":{"denominator":54,"numerator":1,"percentage_points":"+1.85 pp"},"model_configuration":"gemini-3.1-pro-preview_high","paired_original":{"denominator":18,"numerator":17,"percent":"94.44%"},"paired_original_rank":2,"protocol":"mean@1","rank_change":-5},{"clean":{"denominator":1512,"numerator":1453,"percent":"96.10%"},"clean_rank":8,"configuration_id":"gpqa_cfg_v1_c61eab9033f8b5854c5649db1e12ecba8df925dfd4b6c0455cbe678320b35d78","delta":{"denominator":4158,"numerator":137,"percentage_points":"+3.29 pp"},"model_configuration":"gemini-3.5-flash_high","paired_original":{"denominator":264,"numerator":245,"percent":"92.80%"},"paired_original_rank":10,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":189,"numerator":181,"percent":"95.77%"},"clean_rank":9,"configuration_id":"gpqa_cfg_v1_02806112b3fe0147c940db998df739ae579b592620fa54621d718d9ceef20010","delta":{"denominator":4158,"numerator":139,"percentage_points":"+3.34 pp"},"model_configuration":"Grok 4.6 (high)","paired_original":{"denominator":66,"numerator":61,"percent":"92.42%"},"paired_original_rank":12,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":181,"percent":"95.77%"},"clean_rank":9,"configuration_id":"gpqa_cfg_v1_86c15018f283c868987b63ef745efbca663fee38b52cc60a94981c8efd381fab","delta":{"denominator":2079,"numerator":59,"percentage_points":"+2.84 pp"},"model_configuration":"DeepSeek V4 Flash (max)","paired_original":{"denominator":99,"numerator":92,"percent":"92.93%"},"paired_original_rank":8,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":378,"numerator":361,"percent":"95.50%"},"clean_rank":11,"configuration_id":"gpqa_cfg_v1_75b5c82ad8fa3f7520d6c30258334fd2da432a999b7c23be7a10a04ab6bfe1de","delta":{"denominator":33264,"numerator":961,"percentage_points":"+2.89 pp"},"model_configuration":"gemini-3-pro-preview","paired_original":{"denominator":176,"numerator":163,"percent":"92.61%"},"paired_original_rank":11,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":21,"numerator":20,"percent":"95.24%"},"clean_rank":12,"configuration_id":"gpqa_cfg_v1_a30430189e9844037ecd5dd092ea5d76f41c830802306d85f70843d9e087f3cd","delta":{"denominator":462,"numerator":13,"percentage_points":"+2.81 pp"},"model_configuration":"DeepSeek V4 Pro (max)","paired_original":{"denominator":66,"numerator":61,"percent":"92.42%"},"paired_original_rank":12,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":21,"numerator":20,"percent":"95.24%"},"clean_rank":12,"configuration_id":"gpqa_cfg_v1_a4204d1bf0943d754d0da9cb0991f909d6a74aecfad020560a30706bf493ea5d","delta":{"denominator":693,"numerator":16,"percentage_points":"+2.31 pp"},"model_configuration":"claude-opus-5","paired_original":{"denominator":99,"numerator":92,"percent":"92.93%"},"paired_original_rank":8,"protocol":"mean@1","rank_change":-4},{"clean":{"denominator":504,"numerator":479,"percent":"95.04%"},"clean_rank":14,"configuration_id":"gpqa_cfg_v1_0101b0f415adc0308a5aeb576bd938ac45457f4a2b41b2f0ec765daf4da164d7","delta":{"denominator":1848,"numerator":67,"percentage_points":"+3.63 pp"},"model_configuration":"gpt-5.2-2025-12-11_xhigh","paired_original":{"denominator":198,"numerator":181,"percent":"91.41%"},"paired_original_rank":16,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":189,"numerator":179,"percent":"94.71%"},"clean_rank":15,"configuration_id":"gpqa_cfg_v1_211a09311a1e6c34a2d98ce45ee46c8aab95d0380c75c39c262f0d5dd6392e90","delta":{"denominator":4158,"numerator":95,"percentage_points":"+2.28 pp"},"model_configuration":"GPT-5.6 Luna (max)","paired_original":{"denominator":66,"numerator":61,"percent":"92.42%"},"paired_original_rank":12,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":189,"numerator":179,"percent":"94.71%"},"clean_rank":15,"configuration_id":"gpqa_cfg_v1_bc3e18e0e130722c5f1e133ef0ec99c49ebc62c0cf1d9a48f4491777870f83b6","delta":{"denominator":2079,"numerator":79,"percentage_points":"+3.80 pp"},"model_configuration":"qwen3.7-max","paired_original":{"denominator":11,"numerator":10,"percent":"90.91%"},"paired_original_rank":17,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":179,"percent":"94.71%"},"clean_rank":15,"configuration_id":"gpqa_cfg_v1_c1662dd832d707580c960ccdf5da02a7a38f399834c024fd516e6015ff35eaa2","delta":{"denominator":2079,"numerator":58,"percentage_points":"+2.79 pp"},"model_configuration":"kimi-k3_high","paired_original":{"denominator":99,"numerator":91,"percent":"91.92%"},"paired_original_rank":15,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":504,"numerator":473,"percent":"93.85%"},"clean_rank":18,"configuration_id":"gpqa_cfg_v1_74454bd1c2bf40c9cfef8af6222ded9c2c33796a97bb3e76f4227ab403ab23f6","delta":{"denominator":1848,"numerator":59,"percentage_points":"+3.19 pp"},"model_configuration":"gpt-5.5_low","paired_original":{"denominator":396,"numerator":359,"percent":"90.66%"},"paired_original_rank":22,"protocol":"mean@8","rank_change":4},{"clean":{"denominator":81,"numerator":76,"percent":"93.83%"},"clean_rank":19,"configuration_id":"gpqa_cfg_v1_555ebc09735857529c6e51f2975f27f05313997eae5072969000efc2baf8faa7","delta":{"denominator":891,"numerator":26,"percentage_points":"+2.92 pp"},"model_configuration":"kimi-k2.6","paired_original":{"denominator":11,"numerator":10,"percent":"90.91%"},"paired_original_rank":17,"protocol":"mean@6","rank_change":-2},{"clean":{"denominator":63,"numerator":59,"percent":"93.65%"},"clean_rank":20,"configuration_id":"gpqa_cfg_v1_0acf0885980bdc183084c5deb8f6341eda2074c91344c8d89bebce44801f4394","delta":{"denominator":693,"numerator":19,"percentage_points":"+2.74 pp"},"model_configuration":"GPT-5.6 Terra (max)","paired_original":{"denominator":11,"numerator":10,"percent":"90.91%"},"paired_original_rank":17,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":63,"numerator":59,"percent":"93.65%"},"clean_rank":20,"configuration_id":"gpqa_cfg_v1_3209319a9e249f3a03cbe28de9fcd90b9c603d24c23e54f4d19b3acf0eaa894d","delta":{"denominator":693,"numerator":19,"percentage_points":"+2.74 pp"},"model_configuration":"MiniMax-M3","paired_original":{"denominator":11,"numerator":10,"percent":"90.91%"},"paired_original_rank":17,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":63,"numerator":59,"percent":"93.65%"},"clean_rank":20,"configuration_id":"gpqa_cfg_v1_439fcb4bf6820e10b1024c2303f377da9ff1a1106a368e9fbe08acd16c8f32f9","delta":{"denominator":693,"numerator":26,"percentage_points":"+3.75 pp"},"model_configuration":"gpt-5.4-2026-03-05_high","paired_original":{"denominator":99,"numerator":89,"percent":"89.90%"},"paired_original_rank":26,"protocol":"mean@1","rank_change":6},{"clean":{"denominator":63,"numerator":59,"percent":"93.65%"},"clean_rank":20,"configuration_id":"gpqa_cfg_v1_5af397530c003e894fc8b0e8c6dfa9efeb6966457911dae825b08a1198fbb480","delta":{"denominator":154,"numerator":5,"percentage_points":"+3.25 pp"},"model_configuration":"DeepSeek V4 Flash (high)","paired_original":{"denominator":198,"numerator":179,"percent":"90.40%"},"paired_original_rank":24,"protocol":"mean@1","rank_change":4},{"clean":{"denominator":63,"numerator":59,"percent":"93.65%"},"clean_rank":20,"configuration_id":"gpqa_cfg_v1_6fb295e570562ea20a65bcb23229741b5939aa34be0bd746777e7ba4407013f1","delta":{"denominator":2772,"numerator":97,"percentage_points":"+3.50 pp"},"model_configuration":"claude-opus-4-7_xhigh","paired_original":{"denominator":132,"numerator":119,"percent":"90.15%"},"paired_original_rank":25,"protocol":"mean@8","rank_change":5},{"clean":{"denominator":63,"numerator":59,"percent":"93.65%"},"clean_rank":20,"configuration_id":"gpqa_cfg_v1_84d05ad544dcd1793dbdc24f941a24829266c8784cb6aeac8d0b08838b5b63f4","delta":{"denominator":693,"numerator":26,"percentage_points":"+3.75 pp"},"model_configuration":"gpt-5.6-sol_low","paired_original":{"denominator":99,"numerator":89,"percent":"89.90%"},"paired_original_rank":26,"protocol":"mean@1","rank_change":6},{"clean":{"denominator":63,"numerator":59,"percent":"93.65%"},"clean_rank":20,"configuration_id":"gpqa_cfg_v1_d1854f026c5b433ca83c02a7d0582332f2eb4594f2a0f3223a075ac97db14769","delta":{"denominator":5544,"numerator":173,"percentage_points":"+3.12 pp"},"model_configuration":"claude-opus-4-6_32K","paired_original":{"denominator":264,"numerator":239,"percent":"90.53%"},"paired_original_rank":23,"protocol":"mean@8","rank_change":3},{"clean":{"denominator":189,"numerator":176,"percent":"93.12%"},"clean_rank":27,"configuration_id":"gpqa_cfg_v1_17d1971555900db9f02b08fe2063bb81c1f406eb3f99ded57d6366a966344171","delta":{"denominator":2079,"numerator":67,"percentage_points":"+3.22 pp"},"model_configuration":"glm-5.1","paired_original":{"denominator":99,"numerator":89,"percent":"89.90%"},"paired_original_rank":26,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":189,"numerator":176,"percent":"93.12%"},"clean_rank":27,"configuration_id":"gpqa_cfg_v1_e0339f47a0e11c791da4e131cf06ead62b4fe9b676d100b883e92ae30cec47ad","delta":{"denominator":2079,"numerator":46,"percentage_points":"+2.21 pp"},"model_configuration":"deepseek-v4-pro_high","paired_original":{"denominator":11,"numerator":10,"percent":"90.91%"},"paired_original_rank":17,"protocol":"mean@1","rank_change":-10},{"clean":{"denominator":27,"numerator":25,"percent":"92.59%"},"clean_rank":29,"configuration_id":"gpqa_cfg_v1_319d02deeba5eff1fe5e7b0068499b112d9a54efc9456189fc1777dbacdee7a6","delta":{"denominator":27,"numerator":1,"percentage_points":"+3.70 pp"},"model_configuration":"gemini-3.5-flash_low","paired_original":{"denominator":9,"numerator":8,"percent":"88.89%"},"paired_original_rank":30,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":63,"numerator":58,"percent":"92.06%"},"clean_rank":30,"configuration_id":"gpqa_cfg_v1_1e7553c1137d1a896e5336d0f9e36f600eb3f8c1981c9496e39923a6dceac58a","delta":{"denominator":63,"numerator":2,"percentage_points":"+3.17 pp"},"model_configuration":"gpt-5.4-2026-03-05_medium","paired_original":{"denominator":9,"numerator":8,"percent":"88.89%"},"paired_original_rank":30,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":63,"numerator":58,"percent":"92.06%"},"clean_rank":30,"configuration_id":"gpqa_cfg_v1_c09343b2de7a3b30deceaaba8fcf1b1e2930f59ccdd6ba4efcd34c200b253701","delta":{"denominator":1386,"numerator":37,"percentage_points":"+2.67 pp"},"model_configuration":"gemini-3-flash-preview_high","paired_original":{"denominator":66,"numerator":59,"percent":"89.39%"},"paired_original_rank":29,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":1323,"numerator":1213,"percent":"91.69%"},"clean_rank":32,"configuration_id":"gpqa_cfg_v1_4cff4744741c24d1db4b97e733fde765e5f1c2056705a760369c260e9001e3d0","delta":{"denominator":29106,"numerator":919,"percentage_points":"+3.16 pp"},"model_configuration":"claude-opus-4-6_64K","paired_original":{"denominator":462,"numerator":409,"percent":"88.53%"},"paired_original_rank":32,"protocol":"mean@7","rank_change":0},{"clean":{"denominator":189,"numerator":173,"percent":"91.53%"},"clean_rank":33,"configuration_id":"gpqa_cfg_v1_1752768af2b13dfd2f0a9df6802e29b14ab9387775bdd045c8ea1c437f8607f9","delta":{"denominator":4158,"numerator":131,"percentage_points":"+3.15 pp"},"model_configuration":"claude-opus-4-6_max","paired_original":{"denominator":198,"numerator":175,"percent":"88.38%"},"paired_original_rank":33,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":756,"numerator":691,"percent":"91.40%"},"clean_rank":34,"configuration_id":"gpqa_cfg_v1_08b0ee7e2e105ac8e5ed96cd2f47a42dff3bba9cd0b1c886aac128321d37784f","delta":{"denominator":3024,"numerator":97,"percentage_points":"+3.21 pp"},"model_configuration":"gpt-5.2-2025-12-11_high","paired_original":{"denominator":144,"numerator":127,"percent":"88.19%"},"paired_original_rank":36,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":189,"numerator":172,"percent":"91.01%"},"clean_rank":35,"configuration_id":"gpqa_cfg_v1_18c8f9728ba3bccb250dcce8c3a5d52ac4be8aae99733f291cecb08e35b283db","delta":{"denominator":2079,"numerator":65,"percentage_points":"+3.13 pp"},"model_configuration":"claude-opus-5_low","paired_original":{"denominator":33,"numerator":29,"percent":"87.88%"},"paired_original_rank":37,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":172,"percent":"91.01%"},"clean_rank":35,"configuration_id":"gpqa_cfg_v1_538540e64da939fe97f8685591c8f9e03cd907c322ec906eea6256ce46fc72df","delta":{"denominator":2079,"numerator":65,"percentage_points":"+3.13 pp"},"model_configuration":"glm-5","paired_original":{"denominator":33,"numerator":29,"percent":"87.88%"},"paired_original_rank":37,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":172,"percent":"91.01%"},"clean_rank":35,"configuration_id":"gpqa_cfg_v1_6ad1e1d106980a4cd18c39c471f9ddbfbff5865f9045d01824a081231d140126","delta":{"denominator":4158,"numerator":109,"percentage_points":"+2.62 pp"},"model_configuration":"qwen3.6-plus","paired_original":{"denominator":198,"numerator":175,"percent":"88.38%"},"paired_original_rank":33,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":189,"numerator":172,"percent":"91.01%"},"clean_rank":35,"configuration_id":"gpqa_cfg_v1_f9fede76cf79ca860917f59fb5ebf3126982f366fe21b16ef0e22d4d13f4038a","delta":{"denominator":2079,"numerator":65,"percentage_points":"+3.13 pp"},"model_configuration":"kimi-k2.7-code","paired_original":{"denominator":33,"numerator":29,"percent":"87.88%"},"paired_original_rank":37,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":172,"percent":"91.01%"},"clean_rank":35,"configuration_id":"gpqa_cfg_v1_fe902b8590238e8946b66aae41d7ee7406b4975a2264ad5dc8639eef08bc33ac","delta":{"denominator":4158,"numerator":109,"percentage_points":"+2.62 pp"},"model_configuration":"claude-opus-4-8_low","paired_original":{"denominator":198,"numerator":175,"percent":"88.38%"},"paired_original_rank":33,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":1512,"numerator":1375,"percent":"90.94%"},"clean_rank":40,"configuration_id":"gpqa_cfg_v1_2c3d5349208f65704ed39c3181dd248e34283f155f2c5b6d9238c3eebf0c47b3","delta":{"denominator":16632,"numerator":509,"percentage_points":"+3.06 pp"},"model_configuration":"gpt-5.2-2025-12-11_medium","paired_original":{"denominator":33,"numerator":29,"percent":"87.88%"},"paired_original_rank":37,"protocol":"mean@8","rank_change":-3},{"clean":{"denominator":21,"numerator":19,"percent":"90.48%"},"clean_rank":41,"configuration_id":"gpqa_cfg_v1_25b6ff3864c66bd8fca985a9bbb8460bb4e531eb06f3d77a0f494f3b913eb510","delta":{"denominator":11088,"numerator":323,"percentage_points":"+2.91 pp"},"model_configuration":"fireworks/kimi-k2p5","paired_original":{"denominator":1584,"numerator":1387,"percent":"87.56%"},"paired_original_rank":43,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":21,"numerator":19,"percent":"90.48%"},"clean_rank":41,"configuration_id":"gpqa_cfg_v1_5c21ac281f3717fbc022623ee94cf1b26302c0a4bedd910769f1e8033717239a","delta":{"denominator":77,"numerator":2,"percentage_points":"+2.60 pp"},"model_configuration":"qwen3.7-plus","paired_original":{"denominator":33,"numerator":29,"percent":"87.88%"},"paired_original_rank":37,"protocol":"mean@1","rank_change":-4},{"clean":{"denominator":21,"numerator":19,"percent":"90.48%"},"clean_rank":41,"configuration_id":"gpqa_cfg_v1_e3b6650e66d0e9441ad0ed99a22261238f2ac8b4e5ede02f20cde8b6b5dd76c7","delta":{"denominator":1386,"numerator":43,"percentage_points":"+3.10 pp"},"model_configuration":"gpt-5.6-terra_low","paired_original":{"denominator":198,"numerator":173,"percent":"87.37%"},"paired_original_rank":44,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":21,"numerator":19,"percent":"90.48%"},"clean_rank":41,"configuration_id":"gpqa_cfg_v1_fefffa8ea64c405943e964126ee6c767f5126e9879cdf3c2989670b3db7d61ba","delta":{"denominator":1386,"numerator":43,"percentage_points":"+3.10 pp"},"model_configuration":"qwen3.6-max-preview","paired_original":{"denominator":198,"numerator":173,"percent":"87.37%"},"paired_original_rank":44,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":1512,"numerator":1367,"percent":"90.41%"},"clean_rank":45,"configuration_id":"gpqa_cfg_v1_7f8e8f18de3290be2ccfe3701c2d2a4c8a6512f34dc46b7db05c5915a94d40e5","delta":{"denominator":16632,"numerator":463,"percentage_points":"+2.78 pp"},"model_configuration":"gpt-5.1-2025-11-13_high","paired_original":{"denominator":396,"numerator":347,"percent":"87.63%"},"paired_original_rank":42,"protocol":"mean@8","rank_change":-3},{"clean":{"denominator":189,"numerator":170,"percent":"89.95%"},"clean_rank":46,"configuration_id":"gpqa_cfg_v1_5deb548701e55fbdff283a419bd403f4d1d5b178a1239bc33ecc4508382d6b91","delta":{"denominator":4158,"numerator":149,"percentage_points":"+3.58 pp"},"model_configuration":"gemini-3.6-flash_low","paired_original":{"denominator":22,"numerator":19,"percent":"86.36%"},"paired_original_rank":49,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":170,"percent":"89.95%"},"clean_rank":46,"configuration_id":"gpqa_cfg_v1_8db4d324508b283804b4f687b846605f27d9446a2eb2c449127b198717926f5a","delta":{"denominator":2079,"numerator":64,"percentage_points":"+3.08 pp"},"model_configuration":"gpt-5.4-mini-2026-03-17_xhigh","paired_original":{"denominator":99,"numerator":86,"percent":"86.87%"},"paired_original_rank":47,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":170,"percent":"89.95%"},"clean_rank":46,"configuration_id":"gpqa_cfg_v1_a9fa8925e31e2718da34308ac52093f3a67b1df4c9d64b4dd42be93e3f710655","delta":{"denominator":4158,"numerator":149,"percentage_points":"+3.58 pp"},"model_configuration":"gemini-3.5-flash_minimal","paired_original":{"denominator":22,"numerator":19,"percent":"86.36%"},"paired_original_rank":49,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":170,"percent":"89.95%"},"clean_rank":46,"configuration_id":"gpqa_cfg_v1_f657ce567632b0af5c3025bd73318b5a4796ffe43af7c00f6a98baea08eea2b2","delta":{"denominator":2079,"numerator":64,"percentage_points":"+3.08 pp"},"model_configuration":"Grok 4.3 (high)","paired_original":{"denominator":99,"numerator":86,"percent":"86.87%"},"paired_original_rank":47,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":126,"numerator":113,"percent":"89.68%"},"clean_rank":50,"configuration_id":"gpqa_cfg_v1_cf7a0ccdcb575260739a0107fbf38e7b46265e7cca9bb7f750093cf0cbf5cc6c","delta":{"denominator":693,"numerator":16,"percentage_points":"+2.31 pp"},"model_configuration":"claude-sonnet-4-6_32K","paired_original":{"denominator":198,"numerator":173,"percent":"87.37%"},"paired_original_rank":44,"protocol":"mean@8","rank_change":-6},{"clean":{"denominator":756,"numerator":677,"percent":"89.55%"},"clean_rank":51,"configuration_id":"gpqa_cfg_v1_4f3803305451e61c9246b82a829be2d36ae588a015920201efed023d91dabfaa","delta":{"denominator":33264,"numerator":1123,"percentage_points":"+3.38 pp"},"model_configuration":"gpt-5-2025-08-07_high","paired_original":{"denominator":528,"numerator":455,"percent":"86.17%"},"paired_original_rank":53,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":189,"numerator":169,"percent":"89.42%"},"clean_rank":52,"configuration_id":"gpqa_cfg_v1_114b61ba4b46f7fbd32882dba76813956cbdfdc2782903c93f510f22a3b1fc54","delta":{"denominator":4158,"numerator":127,"percentage_points":"+3.05 pp"},"model_configuration":"claude-opus-4-7_max","paired_original":{"denominator":22,"numerator":19,"percent":"86.36%"},"paired_original_rank":49,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":189,"numerator":169,"percent":"89.42%"},"clean_rank":52,"configuration_id":"gpqa_cfg_v1_9fdc200cc70fb609af6cd2ea5c6ec5818e339138000d351fc142b07d30c01378","delta":{"denominator":2079,"numerator":74,"percentage_points":"+3.56 pp"},"model_configuration":"claude-fable-5_max","paired_original":{"denominator":99,"numerator":85,"percent":"85.86%"},"paired_original_rank":55,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":169,"percent":"89.42%"},"clean_rank":52,"configuration_id":"gpqa_cfg_v1_a0a4273f3e80a8760824125f9fe0ff0835270ab6ca17ddf3d6af4f4400e5ecd9","delta":{"denominator":2079,"numerator":74,"percentage_points":"+3.56 pp"},"model_configuration":"qwen3.5-397b-a17b","paired_original":{"denominator":99,"numerator":85,"percent":"85.86%"},"paired_original_rank":55,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":169,"percent":"89.42%"},"clean_rank":52,"configuration_id":"gpqa_cfg_v1_c1890a410e08444aa2e81859f26860a80431a4a5028c50bc7f3514cbc2e50a82","delta":{"denominator":4158,"numerator":127,"percentage_points":"+3.05 pp"},"model_configuration":"qwen3.5-397b-a17b_none","paired_original":{"denominator":22,"numerator":19,"percent":"86.36%"},"paired_original_rank":49,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":9,"numerator":8,"percent":"88.89%"},"clean_rank":56,"configuration_id":"gpqa_cfg_v1_d6f20b74ffa7e4b76d5c03ebd92e8ebb6053e6c47efc1d83f30e9db73129cb4f","delta":{"denominator":33,"numerator":1,"percentage_points":"+3.03 pp"},"model_configuration":"qwen3.6-27b","paired_original":{"denominator":99,"numerator":85,"percent":"85.86%"},"paired_original_rank":55,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":756,"numerator":671,"percent":"88.76%"},"clean_rank":57,"configuration_id":"gpqa_cfg_v1_cad9ce1ea155bb333c5ebd7a19fec521f9db3f4590d2bf1ec498004a3a8f0362","delta":{"denominator":33264,"numerator":901,"percentage_points":"+2.71 pp"},"model_configuration":"claude-opus-4-5-20251101_32K","paired_original":{"denominator":1584,"numerator":1363,"percent":"86.05%"},"paired_original_rank":54,"protocol":"mean@8","rank_change":-3},{"clean":{"denominator":1323,"numerator":1174,"percent":"88.74%"},"clean_rank":58,"configuration_id":"gpqa_cfg_v1_7c3ec07ae3ad1a5f35ccd1d892d17caa2ff48036699a386811c33885de02853a","delta":{"denominator":14553,"numerator":482,"percentage_points":"+3.31 pp"},"model_configuration":"gpt-5-2025-08-07_medium","paired_original":{"denominator":693,"numerator":592,"percent":"85.43%"},"paired_original_rank":60,"protocol":"mean@7","rank_change":2},{"clean":{"denominator":168,"numerator":149,"percent":"88.69%"},"clean_rank":59,"configuration_id":"gpqa_cfg_v1_d7b58af1465d20d5223ea9f8a5e54b764667d1cb51518b107c4dbfe5c939e1fd","delta":{"denominator":2772,"numerator":89,"percentage_points":"+3.21 pp"},"model_configuration":"claude-opus-4-5-20251101_16K","paired_original":{"denominator":792,"numerator":677,"percent":"85.48%"},"paired_original_rank":59,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":189,"numerator":167,"percent":"88.36%"},"clean_rank":60,"configuration_id":"gpqa_cfg_v1_a7d0e1353fd8be2dc557c8c94f0706588043b35fd3151e4c1a3ca31801ecb166","delta":{"denominator":2079,"numerator":52,"percentage_points":"+2.50 pp"},"model_configuration":"gemini-3.6-flash_minimal","paired_original":{"denominator":99,"numerator":85,"percent":"85.86%"},"paired_original_rank":55,"protocol":"mean@1","rank_change":-5},{"clean":{"denominator":189,"numerator":167,"percent":"88.36%"},"clean_rank":60,"configuration_id":"gpqa_cfg_v1_c413680e8fd715c2789e6cbb466901dd7cf9f221fe5a9a6f5aab61e02c530288","delta":{"denominator":2079,"numerator":73,"percentage_points":"+3.51 pp"},"model_configuration":"qwen3.5-plus","paired_original":{"denominator":33,"numerator":28,"percent":"84.85%"},"paired_original_rank":64,"protocol":"mean@1","rank_change":4},{"clean":{"denominator":189,"numerator":167,"percent":"88.36%"},"clean_rank":60,"configuration_id":"gpqa_cfg_v1_d255d05df7d6452741ae49d1c8831a4ac1bdeb8644fc8d5d6b034ded7651c060","delta":{"denominator":2079,"numerator":73,"percentage_points":"+3.51 pp"},"model_configuration":"kimi-k3_low","paired_original":{"denominator":33,"numerator":28,"percent":"84.85%"},"paired_original_rank":64,"protocol":"mean@1","rank_change":4},{"clean":{"denominator":504,"numerator":445,"percent":"88.29%"},"clean_rank":63,"configuration_id":"gpqa_cfg_v1_457d4d9de46a09e8dda7707cbea4e03840a5686a817ef7dbd62b104c41ed8b9e","delta":{"denominator":1232,"numerator":37,"percentage_points":"+3.00 pp"},"model_configuration":"gemini-2.5-pro","paired_original":{"denominator":1584,"numerator":1351,"percent":"85.29%"},"paired_original_rank":62,"protocol":"mean@8","rank_change":-1},{"clean":{"denominator":189,"numerator":166,"percent":"87.83%"},"clean_rank":64,"configuration_id":"gpqa_cfg_v1_2512d5e4965dc519e0cfe48a64b1f7f33961e8ebe6df554e0815a1b4c98f9bd3","delta":{"denominator":2079,"numerator":62,"percentage_points":"+2.98 pp"},"model_configuration":"gemini-2.5-pro-preview-06-05","paired_original":{"denominator":33,"numerator":28,"percent":"84.85%"},"paired_original_rank":64,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":166,"percent":"87.83%"},"clean_rank":64,"configuration_id":"gpqa_cfg_v1_7422854d7e27c6743941108a0a9a0f278c02b92c202c12726abb5c9bc544adf5","delta":{"denominator":4158,"numerator":103,"percentage_points":"+2.48 pp"},"model_configuration":"claude-opus-4-8_none","paired_original":{"denominator":198,"numerator":169,"percent":"85.35%"},"paired_original_rank":61,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":189,"numerator":166,"percent":"87.83%"},"clean_rank":64,"configuration_id":"gpqa_cfg_v1_c11ad0dd40116f32651cfd124cde890429cc3ff2cb1a8b1b529c9d58661f9850","delta":{"denominator":2079,"numerator":62,"percentage_points":"+2.98 pp"},"model_configuration":"qwen3.6-27b_none","paired_original":{"denominator":33,"numerator":28,"percent":"84.85%"},"paired_original_rank":64,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":166,"percent":"87.83%"},"clean_rank":64,"configuration_id":"gpqa_cfg_v1_d03c3f1680526d4f61dc66ec0cfd3f39d637ecad3f65e348f985bb848fc3c8e7","delta":{"denominator":2079,"numerator":62,"percentage_points":"+2.98 pp"},"model_configuration":"gpt-5.4-2026-03-05_low","paired_original":{"denominator":33,"numerator":28,"percent":"84.85%"},"paired_original_rank":64,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":166,"percent":"87.83%"},"clean_rank":64,"configuration_id":"gpqa_cfg_v1_efb979697550c6b7b9d2f7fbf5e13c8cdcdc7f6b1ed66a691f8f08a9078593bc","delta":{"denominator":2079,"numerator":62,"percentage_points":"+2.98 pp"},"model_configuration":"qwen3.6-35b-a3b_none","paired_original":{"denominator":33,"numerator":28,"percent":"84.85%"},"paired_original_rank":64,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":1512,"numerator":1327,"percent":"87.76%"},"clean_rank":69,"configuration_id":"gpqa_cfg_v1_0ec60fbcfbd0a5dc9ad3b1d16e08bf7c3678cacdd15194446fb742b017dc743b","delta":{"denominator":33264,"numerator":907,"percentage_points":"+2.73 pp"},"model_configuration":"gpt-5.1-2025-11-13_medium","paired_original":{"denominator":528,"numerator":449,"percent":"85.04%"},"paired_original_rank":63,"protocol":"mean@8","rank_change":-6},{"clean":{"denominator":63,"numerator":55,"percent":"87.30%"},"clean_rank":70,"configuration_id":"gpqa_cfg_v1_c254d074bf7d4d562416e11363147be58eab58afb4c70c33aee01ccb278b50a8","delta":{"denominator":231,"numerator":8,"percentage_points":"+3.46 pp"},"model_configuration":"gemini-2.5-pro-exp-03-25","paired_original":{"denominator":99,"numerator":83,"percent":"83.84%"},"paired_original_rank":71,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":164,"percent":"86.77%"},"clean_rank":71,"configuration_id":"gpqa_cfg_v1_1df69af1ab0ce9fcc3c822f72011312d0cb297c6ba8c857f0c29740914d904fd","delta":{"denominator":2079,"numerator":61,"percentage_points":"+2.93 pp"},"model_configuration":"qwen3.6-35b-a3b","paired_original":{"denominator":99,"numerator":83,"percent":"83.84%"},"paired_original_rank":71,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":164,"percent":"86.77%"},"clean_rank":71,"configuration_id":"gpqa_cfg_v1_9b2222a171dd46e01575f8e6e673c3b8b1659464ddb17ad4146c5e1da3a26a81","delta":{"denominator":378,"numerator":13,"percentage_points":"+3.44 pp"},"model_configuration":"claude-fable-5_high","paired_original":{"denominator":6,"numerator":5,"percent":"83.33%"},"paired_original_rank":74,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":164,"percent":"86.77%"},"clean_rank":71,"configuration_id":"gpqa_cfg_v1_eb8810fec6811f5a48d750f48d865427a72cac5e58054882006e251bfa0c646c","delta":{"denominator":378,"numerator":13,"percentage_points":"+3.44 pp"},"model_configuration":"claude-sonnet-4-6_high","paired_original":{"denominator":6,"numerator":5,"percent":"83.33%"},"paired_original_rank":74,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":164,"percent":"86.77%"},"clean_rank":71,"configuration_id":"gpqa_cfg_v1_fee207fd49a1b7d70751742c0f126d3d087b989d89849747726ee9468fdf26f8","delta":{"denominator":8316,"numerator":265,"percentage_points":"+3.19 pp"},"model_configuration":"gpt-5.4-mini-2026-03-17_high","paired_original":{"denominator":396,"numerator":331,"percent":"83.59%"},"paired_original_rank":73,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":126,"numerator":109,"percent":"86.51%"},"clean_rank":75,"configuration_id":"gpqa_cfg_v1_f76c85c895ef2ab274d28dcefa237ebea5b991f40fdb01ee93070b5634a7ca73","delta":{"denominator":5544,"numerator":127,"percentage_points":"+2.29 pp"},"model_configuration":"kimi-k2-thinking-turbo","paired_original":{"denominator":792,"numerator":667,"percent":"84.22%"},"paired_original_rank":70,"protocol":"mean@8","rank_change":-5},{"clean":{"denominator":189,"numerator":163,"percent":"86.24%"},"clean_rank":76,"configuration_id":"gpqa_cfg_v1_5e269fbf7b862f78bdfb58b4e2d6a616de3f24a13172a171b7911c19e5882ee3","delta":{"denominator":378,"numerator":11,"percentage_points":"+2.91 pp"},"model_configuration":"claude-sonnet-4-6_medium","paired_original":{"denominator":6,"numerator":5,"percent":"83.33%"},"paired_original_rank":74,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":189,"numerator":163,"percent":"86.24%"},"clean_rank":76,"configuration_id":"gpqa_cfg_v1_ed668418bc3a7abc717016181a36d45c94836c4061801b735c989f31101fc953","delta":{"denominator":2079,"numerator":71,"percentage_points":"+3.42 pp"},"model_configuration":"gpt-5.6-sol_none","paired_original":{"denominator":99,"numerator":82,"percent":"82.83%"},"paired_original_rank":81,"protocol":"mean@1","rank_change":5},{"clean":{"denominator":441,"numerator":380,"percent":"86.17%"},"clean_rank":78,"configuration_id":"gpqa_cfg_v1_b79f40ee05c16a27a3fae7d506fb2e654418d9d40244460a93ad4d4fc6b45468","delta":{"denominator":882,"numerator":25,"percentage_points":"+2.83 pp"},"model_configuration":"deepseek-reasoner","paired_original":{"denominator":6,"numerator":5,"percent":"83.33%"},"paired_original_rank":74,"protocol":"mean@7","rank_change":-4},{"clean":{"denominator":504,"numerator":433,"percent":"85.91%"},"clean_rank":79,"configuration_id":"gpqa_cfg_v1_30fd0582710efe31443fff4d850caba23883e98a0a3df61e9dcbb85e5fb3ad3e","delta":{"denominator":924,"numerator":25,"percentage_points":"+2.71 pp"},"model_configuration":"gemini-3-flash-preview","paired_original":{"denominator":792,"numerator":659,"percent":"83.21%"},"paired_original_rank":80,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":7,"numerator":6,"percent":"85.71%"},"clean_rank":80,"configuration_id":"gpqa_cfg_v1_0bdf7182d5f04c31013cbf0810c5d35c04d75be94de6c6030b1439ce8d8f0278","delta":{"denominator":1386,"numerator":47,"percentage_points":"+3.39 pp"},"model_configuration":"claude-sonnet-4-5-20250929_59K","paired_original":{"denominator":198,"numerator":163,"percent":"82.32%"},"paired_original_rank":83,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":7,"numerator":6,"percent":"85.71%"},"clean_rank":80,"configuration_id":"gpqa_cfg_v1_73606fc2f28e306f4091aca42290cd5715fed6ba0d38ff358cc6a8ec8c2cc294","delta":{"denominator":1386,"numerator":47,"percentage_points":"+3.39 pp"},"model_configuration":"qwen3.5-flash","paired_original":{"denominator":198,"numerator":163,"percent":"82.32%"},"paired_original_rank":83,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":7,"numerator":6,"percent":"85.71%"},"clean_rank":80,"configuration_id":"gpqa_cfg_v1_91d52572fee638e3cdcba8d3d9db001f9c9ce30df32420006ed56c7f4c01fca1","delta":{"denominator":42,"numerator":1,"percentage_points":"+2.38 pp"},"model_configuration":"qwen3.6-flash","paired_original":{"denominator":6,"numerator":5,"percent":"83.33%"},"paired_original_rank":74,"protocol":"mean@1","rank_change":-6},{"clean":{"denominator":7,"numerator":6,"percent":"85.71%"},"clean_rank":80,"configuration_id":"gpqa_cfg_v1_e4fc3d3f5f663e3176f37cb7a7a9414683ad660a66fd87026e1ba802553ac32a","delta":{"denominator":42,"numerator":1,"percentage_points":"+2.38 pp"},"model_configuration":"gemini-3.5-flash-lite_high","paired_original":{"denominator":6,"numerator":5,"percent":"83.33%"},"paired_original_rank":74,"protocol":"mean@1","rank_change":-6},{"clean":{"denominator":216,"numerator":185,"percent":"85.65%"},"clean_rank":84,"configuration_id":"gpqa_cfg_v1_aa564e65d8869f7ebd98b3ce39ff8e2e4e55cefa67b9e71802509a553137b9b3","delta":{"denominator":1188,"numerator":35,"percentage_points":"+2.95 pp"},"model_configuration":"gpt-5.2-2025-12-11_low","paired_original":{"denominator":792,"numerator":655,"percent":"82.70%"},"paired_original_rank":82,"protocol":"mean@8","rank_change":-2},{"clean":{"denominator":27,"numerator":23,"percent":"85.19%"},"clean_rank":85,"configuration_id":"gpqa_cfg_v1_636869f8e677b100d9390e305daa004e4d3cb6ad815d50d572ee8f9ea78baf1f","delta":{"denominator":594,"numerator":17,"percentage_points":"+2.86 pp"},"model_configuration":"gpt-5.6-luna_low","paired_original":{"denominator":198,"numerator":163,"percent":"82.32%"},"paired_original_rank":83,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":27,"numerator":23,"percent":"85.19%"},"clean_rank":85,"configuration_id":"gpqa_cfg_v1_e7c76115b41d855efcf0f7ae19b2b6d7a23a44cfa36c8e7a39886451c18c0663","delta":{"denominator":594,"numerator":17,"percentage_points":"+2.86 pp"},"model_configuration":"qwen3.7-flash","paired_original":{"denominator":198,"numerator":163,"percent":"82.32%"},"paired_original_rank":83,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":756,"numerator":643,"percent":"85.05%"},"clean_rank":87,"configuration_id":"gpqa_cfg_v1_9eb89da0bad3d09e7ddb8a304b1163ece8e34a5dc839f463fab250c970819ca6","delta":{"denominator":8316,"numerator":269,"percentage_points":"+3.23 pp"},"model_configuration":"o3-2025-04-16_high","paired_original":{"denominator":11,"numerator":9,"percent":"81.82%"},"paired_original_rank":87,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":189,"numerator":160,"percent":"84.66%"},"clean_rank":88,"configuration_id":"gpqa_cfg_v1_028330d0cbeadddca05767705c635f8f34aa6f91322bce88dd9bd97968f4a05a","delta":{"denominator":2079,"numerator":59,"percentage_points":"+2.84 pp"},"model_configuration":"claude-sonnet-4-5-20250929_32K","paired_original":{"denominator":11,"numerator":9,"percent":"81.82%"},"paired_original_rank":87,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":189,"numerator":160,"percent":"84.66%"},"clean_rank":88,"configuration_id":"gpqa_cfg_v1_5d0b6fa9beb00bd723082609126a60e2124f2f089224e8935b7733ee6207c2d7","delta":{"denominator":2079,"numerator":59,"percentage_points":"+2.84 pp"},"model_configuration":"gemini-3.1-flash-lite_high","paired_original":{"denominator":11,"numerator":9,"percent":"81.82%"},"paired_original_rank":87,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":63,"numerator":53,"percent":"84.13%"},"clean_rank":90,"configuration_id":"gpqa_cfg_v1_32831c12727d118cdd4d29d933e064c6f5efeb1e888228cb3ccaef5ca5e5a6c5","delta":{"denominator":693,"numerator":23,"percentage_points":"+3.32 pp"},"model_configuration":"qwen3.7-flash_none","paired_original":{"denominator":99,"numerator":80,"percent":"80.81%"},"paired_original_rank":91,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":63,"numerator":53,"percent":"84.13%"},"clean_rank":90,"configuration_id":"gpqa_cfg_v1_8f361ae1265b1142ff8c109e0b1829cdd625d41f3ee9506a6cdcdbab3641ee73","delta":{"denominator":693,"numerator":16,"percentage_points":"+2.31 pp"},"model_configuration":"qwen3.7-plus_none","paired_original":{"denominator":11,"numerator":9,"percent":"81.82%"},"paired_original_rank":87,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":189,"numerator":158,"percent":"83.60%"},"clean_rank":92,"configuration_id":"gpqa_cfg_v1_46e58f3c8d85f65c9905d98e961570ffe8b923db2111db25bcf593e5a586ba5b","delta":{"denominator":4158,"numerator":137,"percentage_points":"+3.29 pp"},"model_configuration":"claude-sonnet-5_max","paired_original":{"denominator":66,"numerator":53,"percent":"80.30%"},"paired_original_rank":94,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":158,"percent":"83.60%"},"clean_rank":92,"configuration_id":"gpqa_cfg_v1_8ce85e5a084acc7fe3dcf3183d33f605314c4189cf5aec161b8576be34f00bd1","delta":{"denominator":2079,"numerator":58,"percentage_points":"+2.79 pp"},"model_configuration":"o3-2025-04-16_medium","paired_original":{"denominator":99,"numerator":80,"percent":"80.81%"},"paired_original_rank":91,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":189,"numerator":157,"percent":"83.07%"},"clean_rank":94,"configuration_id":"gpqa_cfg_v1_8481fd6a05a2aa64f8278e43afe1cb7914779e0b788942a039de1af847ce744b","delta":{"denominator":4158,"numerator":115,"percentage_points":"+2.77 pp"},"model_configuration":"glm-4.7","paired_original":{"denominator":66,"numerator":53,"percent":"80.30%"},"paired_original_rank":94,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":252,"numerator":209,"percent":"82.94%"},"clean_rank":95,"configuration_id":"gpqa_cfg_v1_b8ff63f0f960b4e2f35e0a38f75a361a14b4ce8d403568ee967f98dc1ff60ed7","delta":{"denominator":1232,"numerator":41,"percentage_points":"+3.33 pp"},"model_configuration":"o4-mini-2025-04-16_high","paired_original":{"denominator":1584,"numerator":1261,"percent":"79.61%"},"paired_original_rank":98,"protocol":"mean@8","rank_change":3},{"clean":{"denominator":168,"numerator":139,"percent":"82.74%"},"clean_rank":96,"configuration_id":"gpqa_cfg_v1_9d62b338e275defb1013265362d0e2e54512238ef149f471ef7b2873e01a0b0d","delta":{"denominator":924,"numerator":19,"percentage_points":"+2.06 pp"},"model_configuration":"claude-opus-4-5-20251101","paired_original":{"denominator":88,"numerator":71,"percent":"80.68%"},"paired_original_rank":93,"protocol":"mean@8","rank_change":-3},{"clean":{"denominator":63,"numerator":52,"percent":"82.54%"},"clean_rank":97,"configuration_id":"gpqa_cfg_v1_1ebdf14944b16525b20d369d7989c36385e707ef96468f681ac093d3ef05807e","delta":{"denominator":693,"numerator":19,"percentage_points":"+2.74 pp"},"model_configuration":"o3-2025-04-16_low","paired_original":{"denominator":99,"numerator":79,"percent":"79.80%"},"paired_original_rank":97,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":63,"numerator":52,"percent":"82.54%"},"clean_rank":97,"configuration_id":"gpqa_cfg_v1_7855e86e24fde2ea81d498e4c025c7acf2b0fe6d817cfe394ef7306ce637d247","delta":{"denominator":693,"numerator":26,"percentage_points":"+3.75 pp"},"model_configuration":"claude-sonnet-4-5-20250929_16K","paired_original":{"denominator":33,"numerator":26,"percent":"78.79%"},"paired_original_rank":99,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":378,"numerator":311,"percent":"82.28%"},"clean_rank":99,"configuration_id":"gpqa_cfg_v1_16b158ba0a71b9966a8c0e86927eb1da15d0ee18a198e905d951edda2884f53d","delta":{"denominator":8316,"numerator":185,"percentage_points":"+2.22 pp"},"model_configuration":"Qwen3-235B-A22B-Thinking-2507","paired_original":{"denominator":396,"numerator":317,"percent":"80.05%"},"paired_original_rank":96,"protocol":"mean@2","rank_change":-3},{"clean":{"denominator":189,"numerator":155,"percent":"82.01%"},"clean_rank":100,"configuration_id":"gpqa_cfg_v1_0780c6b4ba641120d8293909ee3b41d284184448300fe82845ccd82ed21a22a8","delta":{"denominator":2079,"numerator":67,"percentage_points":"+3.22 pp"},"model_configuration":"claude-fable-5_low","paired_original":{"denominator":33,"numerator":26,"percent":"78.79%"},"paired_original_rank":99,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":27,"numerator":22,"percent":"81.48%"},"clean_rank":101,"configuration_id":"gpqa_cfg_v1_72074cf78c83a95bdcd075539dbc4c33faf276be230e0adcaf85fd54c8a051ed","delta":{"denominator":297,"numerator":8,"percentage_points":"+2.69 pp"},"model_configuration":"claude-sonnet-4-6_max","paired_original":{"denominator":33,"numerator":26,"percent":"78.79%"},"paired_original_rank":99,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":216,"numerator":175,"percent":"81.02%"},"clean_rank":102,"configuration_id":"gpqa_cfg_v1_850dd081ff0af5e05ae06d189189973570eabc763dcdf7b9c49aeb6b97112d76","delta":{"denominator":432,"numerator":11,"percentage_points":"+2.55 pp"},"model_configuration":"gpt-5.4-nano-2026-03-17_high","paired_original":{"denominator":144,"numerator":113,"percent":"78.47%"},"paired_original_rank":102,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":189,"numerator":152,"percent":"80.42%"},"clean_rank":103,"configuration_id":"gpqa_cfg_v1_24e612fff1c0d93cfbb8e6ac4c6cd800e066c2b55d58c67b0615a42278896a49","delta":{"denominator":4158,"numerator":89,"percentage_points":"+2.14 pp"},"model_configuration":"claude-sonnet-4-20250514_32K","paired_original":{"denominator":198,"numerator":155,"percent":"78.28%"},"paired_original_rank":103,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":152,"percent":"80.42%"},"clean_rank":103,"configuration_id":"gpqa_cfg_v1_67756590ba7c7aa34e9e5076f7e90d0b4251a570d95d5dc77fcb42a6b854f2ee","delta":{"denominator":4158,"numerator":131,"percentage_points":"+3.15 pp"},"model_configuration":"claude-opus-4-1-20250805_16K","paired_original":{"denominator":22,"numerator":17,"percent":"77.27%"},"paired_original_rank":105,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":152,"percent":"80.42%"},"clean_rank":103,"configuration_id":"gpqa_cfg_v1_cc64a12d6bdca099340cd29a26afe83aeffe666f68bc05604bf7f4e7668632cf","delta":{"denominator":189,"numerator":5,"percentage_points":"+2.65 pp"},"model_configuration":"o4-mini-2025-04-16_medium","paired_original":{"denominator":9,"numerator":7,"percent":"77.78%"},"paired_original_rank":104,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":151,"percent":"79.89%"},"clean_rank":106,"configuration_id":"gpqa_cfg_v1_582401c4d6cff8722feac8c0b9c6982da4d7180861f4671dfc33dd6a447680d5","delta":{"denominator":2079,"numerator":65,"percentage_points":"+3.13 pp"},"model_configuration":"claude-3-7-sonnet-20250219_16K","paired_original":{"denominator":99,"numerator":76,"percent":"76.77%"},"paired_original_rank":109,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":151,"percent":"79.89%"},"clean_rank":106,"configuration_id":"gpqa_cfg_v1_84be1c2e599b07f80232212023c305f33061879967f9ca1cd6e6a2aa92d3430f","delta":{"denominator":2079,"numerator":65,"percentage_points":"+3.13 pp"},"model_configuration":"claude-3-7-sonnet-20250219_32K","paired_original":{"denominator":99,"numerator":76,"percent":"76.77%"},"paired_original_rank":109,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":151,"percent":"79.89%"},"clean_rank":106,"configuration_id":"gpqa_cfg_v1_9e05b4b262c02d63871fd36e5097241f13ddb790d81aed45c3cdaa5f0066fe16","delta":{"denominator":2079,"numerator":65,"percentage_points":"+3.13 pp"},"model_configuration":"claude-opus-4-1-20250805_27K","paired_original":{"denominator":99,"numerator":76,"percent":"76.77%"},"paired_original_rank":109,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":189,"numerator":151,"percent":"79.89%"},"clean_rank":106,"configuration_id":"gpqa_cfg_v1_bdda016ac92da795727c89d7eb197d32b9654ce78d9d202bf7a7aae394898d1b","delta":{"denominator":4158,"numerator":109,"percentage_points":"+2.62 pp"},"model_configuration":"gpt-5.5_none","paired_original":{"denominator":22,"numerator":17,"percent":"77.27%"},"paired_original_rank":105,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":504,"numerator":401,"percent":"79.56%"},"clean_rank":110,"configuration_id":"gpqa_cfg_v1_5b98063364ead570d569428619f6b536cda9944d79c01cd3d1325c9b59abb6bb","delta":{"denominator":1848,"numerator":47,"percentage_points":"+2.54 pp"},"model_configuration":"o3-mini-2025-01-31_high","paired_original":{"denominator":396,"numerator":305,"percent":"77.02%"},"paired_original_rank":108,"protocol":"mean@8","rank_change":-2},{"clean":{"denominator":63,"numerator":50,"percent":"79.37%"},"clean_rank":111,"configuration_id":"gpqa_cfg_v1_211e3b3ef5c20703200d0b2a23fabe754e645ec0860ac579ec797da9ffd88edb","delta":{"denominator":77,"numerator":2,"percentage_points":"+2.60 pp"},"model_configuration":"o1-2024-12-17_high","paired_original":{"denominator":99,"numerator":76,"percent":"76.77%"},"paired_original_rank":109,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":63,"numerator":50,"percent":"79.37%"},"clean_rank":111,"configuration_id":"gpqa_cfg_v1_372f86421b339917a5ea87ce041519de28881fa7f9f39ab618a6079c1668516e","delta":{"denominator":1386,"numerator":43,"percentage_points":"+3.10 pp"},"model_configuration":"grok-3-mini-beta_low","paired_original":{"denominator":198,"numerator":151,"percent":"76.26%"},"paired_original_rank":114,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":63,"numerator":50,"percent":"79.37%"},"clean_rank":111,"configuration_id":"gpqa_cfg_v1_7db9c0fa233c93399222a3bfbac6056222f2bcacd6282d5bb91bc0f71c37ae84","delta":{"denominator":1386,"numerator":29,"percentage_points":"+2.09 pp"},"model_configuration":"gpt-5.6-terra_none","paired_original":{"denominator":22,"numerator":17,"percent":"77.27%"},"paired_original_rank":105,"protocol":"mean@1","rank_change":-6},{"clean":{"denominator":189,"numerator":149,"percent":"78.84%"},"clean_rank":114,"configuration_id":"gpqa_cfg_v1_678fc1ebbe85018672c472fa4980a0f10c18c89f0dccfe222b93697861ddf099","delta":{"denominator":33264,"numerator":835,"percentage_points":"+2.51 pp"},"model_configuration":"DeepSeek-R1-0528","paired_original":{"denominator":528,"numerator":403,"percent":"76.33%"},"paired_original_rank":113,"protocol":"mean@8","rank_change":-1},{"clean":{"denominator":189,"numerator":149,"percent":"78.84%"},"clean_rank":114,"configuration_id":"gpqa_cfg_v1_9d9fcd13faf4c34aa3dfac137ba313739953d4e6cb6b34f7004fe8f8a6bc0016","delta":{"denominator":4158,"numerator":107,"percentage_points":"+2.57 pp"},"model_configuration":"claude-opus-4-20250514_16K","paired_original":{"denominator":198,"numerator":151,"percent":"76.26%"},"paired_original_rank":114,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":148,"percent":"78.31%"},"clean_rank":116,"configuration_id":"gpqa_cfg_v1_3b4449cd7456be9a421061071fd80145c7806c116ccfedd101c8f06b7ea16dc5","delta":{"denominator":2079,"numerator":53,"percentage_points":"+2.55 pp"},"model_configuration":"claude-sonnet-4-20250514_16K","paired_original":{"denominator":33,"numerator":25,"percent":"75.76%"},"paired_original_rank":116,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":148,"percent":"78.31%"},"clean_rank":116,"configuration_id":"gpqa_cfg_v1_569cf92af26ce0effa0d7c6f3499ee8a4b59832cf43499b195482e8caf68b0e0","delta":{"denominator":2079,"numerator":53,"percentage_points":"+2.55 pp"},"model_configuration":"openai/gpt-oss-120b_high","paired_original":{"denominator":33,"numerator":25,"percent":"75.76%"},"paired_original_rank":116,"protocol":"mean@2","rank_change":0},{"clean":{"denominator":189,"numerator":148,"percent":"78.31%"},"clean_rank":116,"configuration_id":"gpqa_cfg_v1_5ba80b136d60d34017baf239819ca37835512b9eb474be2e9b599485c1de4c57","delta":{"denominator":2079,"numerator":53,"percentage_points":"+2.55 pp"},"model_configuration":"gemma-4-31b-it_minimal","paired_original":{"denominator":33,"numerator":25,"percent":"75.76%"},"paired_original_rank":116,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":148,"percent":"78.31%"},"clean_rank":116,"configuration_id":"gpqa_cfg_v1_ca66e0d6fa724848f897b1528b7c9822969050042ffe53bc36c3bad551d8bbbe","delta":{"denominator":2079,"numerator":53,"percentage_points":"+2.55 pp"},"model_configuration":"gemini-3.5-flash-lite_low","paired_original":{"denominator":33,"numerator":25,"percent":"75.76%"},"paired_original_rank":116,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":1512,"numerator":1177,"percent":"77.84%"},"clean_rank":120,"configuration_id":"gpqa_cfg_v1_25fa96ba13ac4c3502661f5ee3958b5fead6b8ed9a5a200ed41263fcfa970d6c","delta":{"denominator":1512,"numerator":43,"percentage_points":"+2.84 pp"},"model_configuration":"gpt-5-mini-2025-08-07_high","paired_original":{"denominator":4,"numerator":3,"percent":"75.00%"},"paired_original_rank":122,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":9,"numerator":7,"percent":"77.78%"},"clean_rank":121,"configuration_id":"gpqa_cfg_v1_06fc60637355bfe507b394a67bd3cf0abc7c538efdf32fa6908642999f53ce6d","delta":{"denominator":198,"numerator":5,"percentage_points":"+2.53 pp"},"model_configuration":"o4-mini-2025-04-16_low","paired_original":{"denominator":198,"numerator":149,"percent":"75.25%"},"paired_original_rank":121,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":146,"percent":"77.25%"},"clean_rank":122,"configuration_id":"gpqa_cfg_v1_9176a8be353931cb51348dc081ae7761a85c8a92f75a1939ae96e66f66b67305","delta":{"denominator":2079,"numerator":31,"percentage_points":"+1.49 pp"},"model_configuration":"o1-2024-12-17_medium","paired_original":{"denominator":33,"numerator":25,"percent":"75.76%"},"paired_original_rank":116,"protocol":"mean@1","rank_change":-6},{"clean":{"denominator":189,"numerator":146,"percent":"77.25%"},"clean_rank":122,"configuration_id":"gpqa_cfg_v1_e53c5a14f4aeac84c113c190e76b609ecf71fe6a8a08960e87cf07e538dfe203","delta":{"denominator":2079,"numerator":52,"percentage_points":"+2.50 pp"},"model_configuration":"gpt-5.4-2026-03-05_none","paired_original":{"denominator":99,"numerator":74,"percent":"74.75%"},"paired_original_rank":123,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":145,"percent":"76.72%"},"clean_rank":124,"configuration_id":"gpqa_cfg_v1_2e20ff4f87d82928ed38f791e0d054c52b489cc802e53dcdfc82bf4d51980434","delta":{"denominator":4158,"numerator":103,"percentage_points":"+2.48 pp"},"model_configuration":"o1-2024-12-17_low","paired_original":{"denominator":66,"numerator":49,"percent":"74.24%"},"paired_original_rank":125,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":145,"percent":"76.72%"},"clean_rank":124,"configuration_id":"gpqa_cfg_v1_75c4a32c0454fc968aa549dba870cfd4e1bd5a1f6c823d4e434b9e24f6f77f18","delta":{"denominator":4158,"numerator":103,"percentage_points":"+2.48 pp"},"model_configuration":"gemini-3.1-flash-lite_low","paired_original":{"denominator":66,"numerator":49,"percent":"74.24%"},"paired_original_rank":125,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":145,"percent":"76.72%"},"clean_rank":124,"configuration_id":"gpqa_cfg_v1_eb0badf81e252fdb63ecf80d40bb300712ba693f20e27eab39d39914053768f9","delta":{"denominator":4158,"numerator":103,"percentage_points":"+2.48 pp"},"model_configuration":"gemini-3.5-flash-lite_minimal","paired_original":{"denominator":66,"numerator":49,"percent":"74.24%"},"paired_original_rank":125,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":1512,"numerator":1157,"percent":"76.52%"},"clean_rank":127,"configuration_id":"gpqa_cfg_v1_90e5ce6b4a67252218334c529e05589fdb776c6a026b382161690d249ff27e18","delta":{"denominator":66528,"numerator":1495,"percentage_points":"+2.25 pp"},"model_configuration":"o3-mini-2025-01-31_medium","paired_original":{"denominator":3168,"numerator":2353,"percent":"74.27%"},"paired_original_rank":124,"protocol":"mean@16","rank_change":-3},{"clean":{"denominator":21,"numerator":16,"percent":"76.19%"},"clean_rank":128,"configuration_id":"gpqa_cfg_v1_afeeefbcbb8acb87e92f6e3238d36a6aaacc8dc9149121f1521a3cb957fcba15","delta":{"denominator":1386,"numerator":41,"percentage_points":"+2.96 pp"},"model_configuration":"gpt-5.2-2025-12-11_none","paired_original":{"denominator":198,"numerator":145,"percent":"73.23%"},"paired_original_rank":130,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":143,"percent":"75.66%"},"clean_rank":129,"configuration_id":"gpqa_cfg_v1_758fdeb186e7e3aa8d76746a779e63abb3f3d53282b88caf7802f648df61985a","delta":{"denominator":4158,"numerator":101,"percentage_points":"+2.43 pp"},"model_configuration":"claude-opus-4-1-20250805","paired_original":{"denominator":198,"numerator":145,"percent":"73.23%"},"paired_original_rank":130,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":143,"percent":"75.66%"},"clean_rank":129,"configuration_id":"gpqa_cfg_v1_a385ae2ba85bd7d7184db8aaf5c23614c6f6b5d4da48acbb3e476fa738e7f196","delta":{"denominator":2079,"numerator":40,"percentage_points":"+1.92 pp"},"model_configuration":"gemini-3.1-flash-lite_minimal","paired_original":{"denominator":99,"numerator":73,"percent":"73.74%"},"paired_original_rank":128,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":189,"numerator":143,"percent":"75.66%"},"clean_rank":129,"configuration_id":"gpqa_cfg_v1_bfdeb68407b430f0101f3ac3ac9ff34e4e1bcbc7034bb257d6a50df851148d7e","delta":{"denominator":4158,"numerator":101,"percentage_points":"+2.43 pp"},"model_configuration":"deepseek-v4-pro_none","paired_original":{"denominator":198,"numerator":145,"percent":"73.23%"},"paired_original_rank":130,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":142,"percent":"75.13%"},"clean_rank":132,"configuration_id":"gpqa_cfg_v1_172a5bf1e81d7ea7354117b64e82e5a863ddd3138848c51d12934571427fcfd5","delta":{"denominator":2079,"numerator":29,"percentage_points":"+1.39 pp"},"model_configuration":"claude-sonnet-4-5-20250929","paired_original":{"denominator":99,"numerator":73,"percent":"73.74%"},"paired_original_rank":128,"protocol":"mean@1","rank_change":-4},{"clean":{"denominator":189,"numerator":142,"percent":"75.13%"},"clean_rank":132,"configuration_id":"gpqa_cfg_v1_3c5870b608f9aaf878456e51a25d8aadb6e0d8e9b1e902670d6b72045777d494","delta":{"denominator":16632,"numerator":421,"percentage_points":"+2.53 pp"},"model_configuration":"qwen3-max-2025-09-23","paired_original":{"denominator":792,"numerator":575,"percent":"72.60%"},"paired_original_rank":133,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":63,"numerator":47,"percent":"74.60%"},"clean_rank":134,"configuration_id":"gpqa_cfg_v1_2d16e5d29fe93a718329e4d401a08f5d17e483de216571aa6a74b0aa9f945c11","delta":{"denominator":693,"numerator":20,"percentage_points":"+2.89 pp"},"model_configuration":"gpt-5-2025-08-07_minimal","paired_original":{"denominator":99,"numerator":71,"percent":"71.72%"},"paired_original_rank":135,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":216,"numerator":161,"percent":"74.54%"},"clean_rank":135,"configuration_id":"gpqa_cfg_v1_00b750e02386f4541fe325efb0a8449df2b72a52b8dd40a582a7a752470e3a3d","delta":{"denominator":4752,"numerator":137,"percentage_points":"+2.88 pp"},"model_configuration":"gpt-5-mini-2025-08-07_medium","paired_original":{"denominator":1584,"numerator":1135,"percent":"71.65%"},"paired_original_rank":137,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":27,"numerator":20,"percent":"74.07%"},"clean_rank":136,"configuration_id":"gpqa_cfg_v1_53004f5f1c902a577afd76e4f9016de7963f5d869c85bb3d7906463afa273244","delta":{"denominator":54,"numerator":1,"percentage_points":"+1.85 pp"},"model_configuration":"gpt-5.4-nano-2026-03-17_low","paired_original":{"denominator":18,"numerator":13,"percent":"72.22%"},"paired_original_rank":134,"protocol":"mean@1","rank_change":-2},{"clean":{"denominator":27,"numerator":20,"percent":"74.07%"},"clean_rank":136,"configuration_id":"gpqa_cfg_v1_e8b7f10a34ef4830b8a98b69d984ed96a9afc3e9e8a7858993bef4958cb0eecc","delta":{"denominator":297,"numerator":7,"percentage_points":"+2.36 pp"},"model_configuration":"gpt-5-mini-2025-08-07_minimal","paired_original":{"denominator":99,"numerator":71,"percent":"71.72%"},"paired_original_rank":135,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":189,"numerator":139,"percent":"73.54%"},"clean_rank":138,"configuration_id":"gpqa_cfg_v1_4c3a34cfb53d1bdeb2350a0a8d43d7dd34e2878b1834dd8a615a4f6100909a4c","delta":{"denominator":4158,"numerator":97,"percentage_points":"+2.33 pp"},"model_configuration":"deepseek-chat","paired_original":{"denominator":66,"numerator":47,"percent":"71.21%"},"paired_original_rank":138,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":139,"percent":"73.54%"},"clean_rank":138,"configuration_id":"gpqa_cfg_v1_b25fbe60dede1fb3425e3063a2463d869b9c4c00d4b6b0aed6903c5818c3f024","delta":{"denominator":4158,"numerator":97,"percentage_points":"+2.33 pp"},"model_configuration":"claude-haiku-4-5-20251001_32K","paired_original":{"denominator":66,"numerator":47,"percent":"71.21%"},"paired_original_rank":138,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":1512,"numerator":1109,"percent":"73.35%"},"clean_rank":140,"configuration_id":"gpqa_cfg_v1_b6e973f9d6f6139a95d79c29885fb8d73070eadc9565d29c3978bdf06333826d","delta":{"denominator":16632,"numerator":439,"percentage_points":"+2.64 pp"},"model_configuration":"qwen3-235b-a22b","paired_original":{"denominator":99,"numerator":70,"percent":"70.71%"},"paired_original_rank":141,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":189,"numerator":137,"percent":"72.49%"},"clean_rank":141,"configuration_id":"gpqa_cfg_v1_8d361a9663df4edd4bbe91b30f4d62bbd4be4a7a778484c5a867483645d19a6c","delta":{"denominator":4158,"numerator":53,"percentage_points":"+1.27 pp"},"model_configuration":"glm-5.2_none","paired_original":{"denominator":66,"numerator":47,"percent":"71.21%"},"paired_original_rank":138,"protocol":"mean@1","rank_change":-3},{"clean":{"denominator":1512,"numerator":1091,"percent":"72.16%"},"clean_rank":142,"configuration_id":"gpqa_cfg_v1_153bcc443e2c2d97f408c94b5455fe8eded12a0a5d244c70e8fe4d7e9700b5e0","delta":{"denominator":1512,"numerator":41,"percentage_points":"+2.71 pp"},"model_configuration":"gpt-5-nano-2025-08-07_high","paired_original":{"denominator":36,"numerator":25,"percent":"69.44%"},"paired_original_rank":142,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":7,"numerator":5,"percent":"71.43%"},"clean_rank":143,"configuration_id":"gpqa_cfg_v1_90806d19c82368364d7c0c55ac1dcd465817154c590e19fccb69699999696009","delta":{"denominator":1386,"numerator":31,"percentage_points":"+2.24 pp"},"model_configuration":"claude-opus-4-20250514","paired_original":{"denominator":198,"numerator":137,"percent":"69.19%"},"paired_original_rank":143,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":134,"percent":"70.90%"},"clean_rank":144,"configuration_id":"gpqa_cfg_v1_68f4f26b76cc464a36f5d412acf484bac54ad4002ceef770bb5a16e84938c442","delta":{"denominator":4158,"numerator":113,"percentage_points":"+2.72 pp"},"model_configuration":"o3-mini-2025-01-31_low","paired_original":{"denominator":22,"numerator":15,"percent":"68.18%"},"paired_original_rank":145,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":189,"numerator":134,"percent":"70.90%"},"clean_rank":144,"configuration_id":"gpqa_cfg_v1_788a0f0a26e95b58f7bad9146781e148246e45068dfba74be09ffb9e967a5f8a","delta":{"denominator":2079,"numerator":46,"percentage_points":"+2.21 pp"},"model_configuration":"gpt-4.5-preview-2025-02-27","paired_original":{"denominator":99,"numerator":68,"percent":"68.69%"},"paired_original_rank":144,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":1512,"numerator":1061,"percent":"70.17%"},"clean_rank":146,"configuration_id":"gpqa_cfg_v1_cc760f480aef1f4d05a51dc48fb94f9bd8123b5a19ed8a8b623a9aa5936069ff","delta":{"denominator":16632,"numerator":457,"percentage_points":"+2.75 pp"},"model_configuration":"gpt-5-nano-2025-08-07_medium","paired_original":{"denominator":132,"numerator":89,"percent":"67.42%"},"paired_original_rank":147,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":56,"numerator":39,"percent":"69.64%"},"clean_rank":147,"configuration_id":"gpqa_cfg_v1_8377771cdc0ba4e2788fa70543be3075e24b56931201a00dc7de01ffaebc6583","delta":{"denominator":1232,"numerator":25,"percentage_points":"+2.03 pp"},"model_configuration":"DeepSeek-V3-0324","paired_original":{"denominator":176,"numerator":119,"percent":"67.61%"},"paired_original_rank":146,"protocol":"mean@8","rank_change":-1},{"clean":{"denominator":189,"numerator":131,"percent":"69.31%"},"clean_rank":148,"configuration_id":"gpqa_cfg_v1_934610fb54d8c86a7bea8d514dee077ee2eb5884977abbe065c70009f334e3b7","delta":{"denominator":8316,"numerator":199,"percentage_points":"+2.39 pp"},"model_configuration":"gpt-4.1-2025-04-14","paired_original":{"denominator":396,"numerator":265,"percent":"66.92%"},"paired_original_rank":149,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":189,"numerator":131,"percent":"69.31%"},"clean_rank":148,"configuration_id":"gpqa_cfg_v1_ebe93e140bced2f2c9f1711124393f7b59a700e61071b4652a4827cd3e713257","delta":{"denominator":189,"numerator":5,"percentage_points":"+2.65 pp"},"model_configuration":"gpt-5.1-2025-11-13_none","paired_original":{"denominator":3,"numerator":2,"percent":"66.67%"},"paired_original_rank":150,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":189,"numerator":130,"percent":"68.78%"},"clean_rank":150,"configuration_id":"gpqa_cfg_v1_54aba346a03524e84a582d320e68f5d6e7e1ca6e1ff7a164ae903b0be1d435a6","delta":{"denominator":189,"numerator":4,"percentage_points":"+2.12 pp"},"model_configuration":"claude-sonnet-4-20250514","paired_original":{"denominator":3,"numerator":2,"percent":"66.67%"},"paired_original_rank":150,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":130,"percent":"68.78%"},"clean_rank":150,"configuration_id":"gpqa_cfg_v1_e965137fed2c696942a43e020de955f796ca37ab01b735427391d3454e309d79","delta":{"denominator":189,"numerator":4,"percentage_points":"+2.12 pp"},"model_configuration":"gemini-2.5-pro-preview-05-06","paired_original":{"denominator":3,"numerator":2,"percent":"66.67%"},"paired_original_rank":150,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":63,"numerator":43,"percent":"68.25%"},"clean_rank":152,"configuration_id":"gpqa_cfg_v1_f1d46d247ebf82560a68e3bc5cf4e07c002603eaadfa51893bf7ed84564bbdbf","delta":{"denominator":3696,"numerator":47,"percentage_points":"+1.27 pp"},"model_configuration":"Llama-4-Maverick-17B-128E-Instruct-FP8","paired_original":{"denominator":1584,"numerator":1061,"percent":"66.98%"},"paired_original_rank":148,"protocol":"mean@8","rank_change":-4},{"clean":{"denominator":1512,"numerator":1031,"percent":"68.19%"},"clean_rank":153,"configuration_id":"gpqa_cfg_v1_71cf41df6359bc5fad4222b789a003c9cc8bf989e0345e4db04c1caeaf870496","delta":{"denominator":33264,"numerator":779,"percentage_points":"+2.34 pp"},"model_configuration":"gpt-4.1-mini-2025-04-14","paired_original":{"denominator":1584,"numerator":1043,"percent":"65.85%"},"paired_original_rank":154,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":1512,"numerator":1031,"percent":"68.19%"},"clean_rank":153,"configuration_id":"gpqa_cfg_v1_cf79ba0cd7bfae3b0d2ce505ab46e7b0b3957b3840464c52248da9bfb033e3d1","delta":{"denominator":8316,"numerator":179,"percentage_points":"+2.15 pp"},"model_configuration":"claude-3-7-sonnet-20250219","paired_original":{"denominator":792,"numerator":523,"percent":"66.04%"},"paired_original_rank":153,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":168,"numerator":113,"percent":"67.26%"},"clean_rank":155,"configuration_id":"gpqa_cfg_v1_4e08dd7d1b1197d17bb458f074103b518735398f7272662a3a46c37f381e18c3","delta":{"denominator":5544,"numerator":103,"percentage_points":"+1.86 pp"},"model_configuration":"qwq-plus","paired_original":{"denominator":396,"numerator":259,"percent":"65.40%"},"paired_original_rank":156,"protocol":"mean@8","rank_change":1},{"clean":{"denominator":189,"numerator":127,"percent":"67.20%"},"clean_rank":156,"configuration_id":"gpqa_cfg_v1_8d5c9aff7847b4d80b03824f778e301bb2f71d82a7d54bc8752a32b2d93880b9","delta":{"denominator":2079,"numerator":32,"percentage_points":"+1.54 pp"},"model_configuration":"gemini-2.0-pro-exp-02-05","paired_original":{"denominator":99,"numerator":65,"percent":"65.66%"},"paired_original_rank":155,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":189,"numerator":125,"percent":"66.14%"},"clean_rank":157,"configuration_id":"gpqa_cfg_v1_6c8c1dd0bd4400f5eb27a320a475692af2d68a39af98d4f0b552fc0419ecb544","delta":{"denominator":4158,"numerator":83,"percentage_points":"+2.00 pp"},"model_configuration":"gpt-5.4-mini-2026-03-17_none","paired_original":{"denominator":198,"numerator":127,"percent":"64.14%"},"paired_original_rank":157,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":189,"numerator":124,"percent":"65.61%"},"clean_rank":158,"configuration_id":"gpqa_cfg_v1_af5c54e09d41d150ba0a812d4c0d2dd9ded66334836f64da4a4bb102999428cd","delta":{"denominator":2079,"numerator":41,"percentage_points":"+1.97 pp"},"model_configuration":"gpt-5.6-luna_none","paired_original":{"denominator":11,"numerator":7,"percent":"63.64%"},"paired_original_rank":159,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":1512,"numerator":991,"percent":"65.54%"},"clean_rank":159,"configuration_id":"gpqa_cfg_v1_e306757163f97ebb8a7e53e35a5f2c8b98039f375eba0fff85b09abd20803643","delta":{"denominator":16632,"numerator":233,"percentage_points":"+1.40 pp"},"model_configuration":"gemini-2.0-flash-001","paired_original":{"denominator":198,"numerator":127,"percent":"64.14%"},"paired_original_rank":157,"protocol":"mean@16","rank_change":-2},{"clean":{"denominator":14,"numerator":9,"percent":"64.29%"},"clean_rank":160,"configuration_id":"gpqa_cfg_v1_54e29f3aa006e93a5e11a4220d4ffe0398120b3ceb40328c11661e50953226b2","delta":{"denominator":2772,"numerator":53,"percentage_points":"+1.91 pp"},"model_configuration":"o1-mini-2024-09-12_high","paired_original":{"denominator":396,"numerator":247,"percent":"62.37%"},"paired_original_rank":160,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":756,"numerator":467,"percent":"61.77%"},"clean_rank":161,"configuration_id":"gpqa_cfg_v1_066a71969e005dcbca3c374019b97a0455b621a9da160724d16808d4ee7bdd9e","delta":{"denominator":16632,"numerator":215,"percentage_points":"+1.29 pp"},"model_configuration":"claude-haiku-4-5-20251001","paired_original":{"denominator":792,"numerator":479,"percent":"60.48%"},"paired_original_rank":161,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":378,"numerator":233,"percent":"61.64%"},"clean_rank":162,"configuration_id":"gpqa_cfg_v1_f77deb38bca0cfa2e68170980f653b63a4bea919a4a2c2475cdc2a37c7b82e48","delta":{"denominator":66528,"numerator":1423,"percentage_points":"+2.14 pp"},"model_configuration":"o1-mini-2024-09-12_medium","paired_original":{"denominator":3168,"numerator":1885,"percent":"59.50%"},"paired_original_rank":163,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":756,"numerator":461,"percent":"60.98%"},"clean_rank":163,"configuration_id":"gpqa_cfg_v1_202023fa5943ac9a7c62a2c1d63c955492f2ac82a4850bb837b96a4adefe8dbe","delta":{"denominator":33264,"numerator":481,"percentage_points":"+1.45 pp"},"model_configuration":"mistral-medium-2505","paired_original":{"denominator":1584,"numerator":943,"percent":"59.53%"},"paired_original_rank":162,"protocol":"mean@8","rank_change":-1},{"clean":{"denominator":27,"numerator":16,"percent":"59.26%"},"clean_rank":164,"configuration_id":"gpqa_cfg_v1_69e0a76e213450c4c99586311e06c32b0188a7d080456a8a8e66f8c8bfd84fdb","delta":{"denominator":297,"numerator":5,"percentage_points":"+1.68 pp"},"model_configuration":"gpt-5-nano-2025-08-07_low","paired_original":{"denominator":33,"numerator":19,"percent":"57.58%"},"paired_original_rank":164,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":27,"numerator":16,"percent":"59.26%"},"clean_rank":164,"configuration_id":"gpqa_cfg_v1_c7c57d8f54855969165427be088ee517a83a6182a8dfb2e215a3d8ce25b8e0ef","delta":{"denominator":594,"numerator":13,"percentage_points":"+2.19 pp"},"model_configuration":"gemini-2.0-flash-thinking-exp-01-21","paired_original":{"denominator":198,"numerator":113,"percent":"57.07%"},"paired_original_rank":166,"protocol":"mean@1","rank_change":2},{"clean":{"denominator":336,"numerator":197,"percent":"58.63%"},"clean_rank":166,"configuration_id":"gpqa_cfg_v1_1d5d4f300a0ef5bd1a8578e997798bae60cb1a0710e2c0c8a37c01ddcaec92c5","delta":{"denominator":22176,"numerator":311,"percentage_points":"+1.40 pp"},"model_configuration":"gemini-1.5-pro-002","paired_original":{"denominator":3168,"numerator":1813,"percent":"57.23%"},"paired_original_rank":165,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":1008,"numerator":583,"percent":"57.84%"},"clean_rank":167,"configuration_id":"gpqa_cfg_v1_2fd227a793f3f537502469f4573a2c42c5e81c9dddbb46cd8400afad108ce3ec","delta":{"denominator":22176,"numerator":289,"percentage_points":"+1.30 pp"},"model_configuration":"DeepSeek-V3","paired_original":{"denominator":352,"numerator":199,"percent":"56.53%"},"paired_original_rank":167,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":756,"numerator":437,"percent":"57.80%"},"clean_rank":168,"configuration_id":"gpqa_cfg_v1_d83a57d702d18c236094a9c8c5928b5c1aee75f30eb5fa16b86d87b7c6f6c312","delta":{"denominator":33264,"numerator":685,"percentage_points":"+2.06 pp"},"model_configuration":"DeepSeek-R1-Distill-Llama-70B","paired_original":{"denominator":1584,"numerator":883,"percent":"55.74%"},"paired_original_rank":171,"protocol":"mean@8","rank_change":3},{"clean":{"denominator":7,"numerator":4,"percent":"57.14%"},"clean_rank":169,"configuration_id":"gpqa_cfg_v1_be00ebd659f5798739f50eb4ddb03eec87164876f7bc5f04ab7cadb78d6ba39c","delta":{"denominator":462,"numerator":5,"percentage_points":"+1.08 pp"},"model_configuration":"magistral-small-2506","paired_original":{"denominator":66,"numerator":37,"percent":"56.06%"},"paired_original_rank":169,"protocol":"mean@1","rank_change":0},{"clean":{"denominator":7,"numerator":4,"percent":"57.14%"},"clean_rank":169,"configuration_id":"gpqa_cfg_v1_de907b3d30ffa90340ef7c1d7ab47a458ec5f37194fde893d5354bb08a8c14fc","delta":{"denominator":63,"numerator":1,"percentage_points":"+1.59 pp"},"model_configuration":"gpt-5.4-nano-2026-03-17_none","paired_original":{"denominator":9,"numerator":5,"percent":"55.56%"},"paired_original_rank":172,"protocol":"mean@1","rank_change":3},{"clean":{"denominator":1512,"numerator":863,"percent":"57.08%"},"clean_rank":171,"configuration_id":"gpqa_cfg_v1_af55e40ea3129f1d80b5438a797e0c4771653bc22e47f0d34ca14dc79968ea23","delta":{"denominator":33264,"numerator":317,"percentage_points":"+0.95 pp"},"model_configuration":"qwen-max-2025-01-25","paired_original":{"denominator":1584,"numerator":889,"percent":"56.12%"},"paired_original_rank":168,"protocol":"mean@8","rank_change":-3},{"clean":{"denominator":72,"numerator":41,"percent":"56.94%"},"clean_rank":172,"configuration_id":"gpqa_cfg_v1_a95dfdcebc4badfda7cb23fb421a952fccefaeb6b649462fb1643f6e2ab604fc","delta":{"denominator":792,"numerator":7,"percentage_points":"+0.88 pp"},"model_configuration":"phi-4","paired_original":{"denominator":66,"numerator":37,"percent":"56.06%"},"paired_original_rank":169,"protocol":"mean@16","rank_change":-3},{"clean":{"denominator":1008,"numerator":569,"percent":"56.45%"},"clean_rank":173,"configuration_id":"gpqa_cfg_v1_d1221f2c8514da265ead360bb375cb64430d6cd83d164f77739373eae7f37879","delta":{"denominator":11088,"numerator":127,"percentage_points":"+1.15 pp"},"model_configuration":"claude-3-5-sonnet-20241022","paired_original":{"denominator":132,"numerator":73,"percent":"55.30%"},"paired_original_rank":173,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":756,"numerator":421,"percent":"55.69%"},"clean_rank":174,"configuration_id":"gpqa_cfg_v1_4cdcf5e7b5a7d772f20148bb6cb03e8e5e57acbe5490e9963e0df1ff0a13f315","delta":{"denominator":8316,"numerator":137,"percentage_points":"+1.65 pp"},"model_configuration":"claude-3-5-sonnet-20240620","paired_original":{"denominator":198,"numerator":107,"percent":"54.04%"},"paired_original_rank":174,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":1008,"numerator":545,"percent":"54.07%"},"clean_rank":175,"configuration_id":"gpqa_cfg_v1_58b6db074aeb963d09941f6ead70727311ab03f1d3e3eb559008d9b988dccb25","delta":{"denominator":11088,"numerator":31,"percentage_points":"+0.28 pp"},"model_configuration":"grok-2-1212","paired_original":{"denominator":132,"numerator":71,"percent":"53.79%"},"paired_original_rank":175,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":378,"numerator":199,"percent":"52.65%"},"clean_rank":176,"configuration_id":"gpqa_cfg_v1_5308e01f281df009ce28185a83659d0ad19a59cdd3e55b5038c6f0b95a53d8b5","delta":{"denominator":33264,"numerator":271,"percentage_points":"+0.81 pp"},"model_configuration":"Llama-4-Scout-17B-16E-Instruct","paired_original":{"denominator":1584,"numerator":821,"percent":"51.83%"},"paired_original_rank":176,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":432,"numerator":227,"percent":"52.55%"},"clean_rank":177,"configuration_id":"gpqa_cfg_v1_902ec55e0bbd019f00882dc1d375e59dbbf9bd5154a35f0daad1099813288565","delta":{"denominator":2376,"numerator":53,"percentage_points":"+2.23 pp"},"model_configuration":"o1-preview-2024-09-12","paired_original":{"denominator":1584,"numerator":797,"percent":"50.32%"},"paired_original_rank":179,"protocol":"mean@16","rank_change":2},{"clean":{"denominator":756,"numerator":397,"percent":"52.51%"},"clean_rank":178,"configuration_id":"gpqa_cfg_v1_c6f1b461149c22be3ff762bed74ced5bc575c8324790a49a899f7f672ee32cfa","delta":{"denominator":66528,"numerator":1063,"percentage_points":"+1.60 pp"},"model_configuration":"Llama-3.1-405B-Instruct","paired_original":{"denominator":3168,"numerator":1613,"percent":"50.92%"},"paired_original_rank":178,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":252,"numerator":131,"percent":"51.98%"},"clean_rank":179,"configuration_id":"gpqa_cfg_v1_41455c97d103f4b57508177da6055ed522be7154caac525ae1d06daba079907f","delta":{"denominator":11088,"numerator":73,"percentage_points":"+0.66 pp"},"model_configuration":"mistral-large-2411","paired_original":{"denominator":528,"numerator":271,"percent":"51.33%"},"paired_original_rank":177,"protocol":"mean@8","rank_change":-2},{"clean":{"denominator":216,"numerator":109,"percent":"50.46%"},"clean_rank":180,"configuration_id":"gpqa_cfg_v1_ee7683485330a0bb484846eff9b707715b9688adb7f1e75b90ec838b53083c08","delta":{"denominator":9504,"numerator":119,"percentage_points":"+1.25 pp"},"model_configuration":"gpt-4o-2024-08-06","paired_original":{"denominator":3168,"numerator":1559,"percent":"49.21%"},"paired_original_rank":180,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":1513,"percent":"50.03%"},"clean_rank":181,"configuration_id":"gpqa_cfg_v1_b36681f056f84dece01834a86b30fd4e56b779b993f6548b69eb449624c0d284","delta":{"denominator":66528,"numerator":589,"percentage_points":"+0.89 pp"},"model_configuration":"qwen2.5-72b-instruct","paired_original":{"denominator":352,"numerator":173,"percent":"49.15%"},"paired_original_rank":181,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":1511,"percent":"49.97%"},"clean_rank":182,"configuration_id":"gpqa_cfg_v1_5169bc0c733d68b48affc9e28c896dd620d7ff8a7a428662b2f13213317c8364","delta":{"denominator":66528,"numerator":713,"percentage_points":"+1.07 pp"},"model_configuration":"gpt-4o-2024-05-13","paired_original":{"denominator":3168,"numerator":1549,"percent":"48.90%"},"paired_original_rank":184,"protocol":"mean@16","rank_change":2},{"clean":{"denominator":1512,"numerator":755,"percent":"49.93%"},"clean_rank":183,"configuration_id":"gpqa_cfg_v1_340eb8d0df205fcfca78e8fe1bef9408f9a08baa2fbe98a058f748c1dce19c09","delta":{"denominator":33264,"numerator":335,"percentage_points":"+1.01 pp"},"model_configuration":"gpt-4.1-nano-2025-04-14","paired_original":{"denominator":1584,"numerator":775,"percent":"48.93%"},"paired_original_rank":183,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":504,"numerator":251,"percent":"49.80%"},"clean_rank":184,"configuration_id":"gpqa_cfg_v1_e4ead69d7e02fd150d9663a883879015c0849b66ff61f05140cf8318e45c4f6d","delta":{"denominator":22176,"numerator":173,"percentage_points":"+0.78 pp"},"model_configuration":"mistral-large-2407","paired_original":{"denominator":3168,"numerator":1553,"percent":"49.02%"},"paired_original_rank":182,"protocol":"mean@16","rank_change":-2},{"clean":{"denominator":1512,"numerator":743,"percent":"49.14%"},"clean_rank":185,"configuration_id":"gpqa_cfg_v1_766ddf5c78cac9398614df89b5abd827a064cc4b8b5645bfcd7cbdf12dbc9cc8","delta":{"denominator":66528,"numerator":835,"percentage_points":"+1.26 pp"},"model_configuration":"gpt-4o-2024-11-20","paired_original":{"denominator":3168,"numerator":1517,"percent":"47.89%"},"paired_original_rank":187,"protocol":"mean@16","rank_change":2},{"clean":{"denominator":189,"numerator":92,"percent":"48.68%"},"clean_rank":186,"configuration_id":"gpqa_cfg_v1_e303a40eed6943f0cc681f7368cff1f8202b2615ff9e972a0f6b2ec5c323afbc","delta":{"denominator":2079,"numerator":4,"percentage_points":"+0.19 pp"},"model_configuration":"gpt-5-nano-2025-08-07_minimal","paired_original":{"denominator":33,"numerator":16,"percent":"48.48%"},"paired_original_rank":185,"protocol":"mean@1","rank_change":-1},{"clean":{"denominator":1512,"numerator":731,"percent":"48.35%"},"clean_rank":187,"configuration_id":"gpqa_cfg_v1_5b80556138a24f63b8f2cb80dccbea6656b707f83aedcd63a5cfe0c1d618c3d1","delta":{"denominator":2079,"numerator":5,"percentage_points":"+0.24 pp"},"model_configuration":"qwen-plus-2025-01-25","paired_original":{"denominator":264,"numerator":127,"percent":"48.11%"},"paired_original_rank":186,"protocol":"mean@8","rank_change":-1},{"clean":{"denominator":27,"numerator":13,"percent":"48.15%"},"clean_rank":188,"configuration_id":"gpqa_cfg_v1_346701427f0d44ced8ca606e9855b570bb07e6ef55855f0a6332eed79ea25189","delta":{"denominator":9504,"numerator":67,"percentage_points":"+0.70 pp"},"model_configuration":"Llama-3.3-70B-Instruct","paired_original":{"denominator":352,"numerator":167,"percent":"47.44%"},"paired_original_rank":189,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":27,"numerator":13,"percent":"48.15%"},"clean_rank":188,"configuration_id":"gpqa_cfg_v1_a0e72767f9478bbe4b04b9a3bea317ca374e27b8e4bbf00393cac2b87d1a2775","delta":{"denominator":594,"numerator":13,"percentage_points":"+2.19 pp"},"model_configuration":"gpt-oss-20b_high","paired_original":{"denominator":198,"numerator":91,"percent":"45.96%"},"paired_original_rank":195,"protocol":"mean@1","rank_change":7},{"clean":{"denominator":3024,"numerator":1453,"percent":"48.05%"},"clean_rank":190,"configuration_id":"gpqa_cfg_v1_a655fc2ed8e0041b02a2ea22c580e50c6da8f701a9667563da361120fddcdab2","delta":{"denominator":66528,"numerator":487,"percentage_points":"+0.73 pp"},"model_configuration":"gemini-1.5-flash-002","paired_original":{"denominator":3168,"numerator":1499,"percent":"47.32%"},"paired_original_rank":190,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":252,"numerator":121,"percent":"48.02%"},"clean_rank":191,"configuration_id":"gpqa_cfg_v1_fe008500b306f51fcb6fd66c7d472ecf21625ed009686be7a81709ebd6aad0c7","delta":{"denominator":924,"numerator":5,"percentage_points":"+0.54 pp"},"model_configuration":"mistral-small-2503","paired_original":{"denominator":99,"numerator":47,"percent":"47.47%"},"paired_original_rank":188,"protocol":"mean@8","rank_change":-3},{"clean":{"denominator":378,"numerator":181,"percent":"47.88%"},"clean_rank":192,"configuration_id":"gpqa_cfg_v1_59886bda1c8504af950b9962ec90464beff5b80e797a4d747fee5b15546d0674","delta":{"denominator":33264,"numerator":241,"percentage_points":"+0.72 pp"},"model_configuration":"claude-3-opus-20240229","paired_original":{"denominator":176,"numerator":83,"percent":"47.16%"},"paired_original_rank":191,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":756,"numerator":355,"percent":"46.96%"},"clean_rank":193,"configuration_id":"gpqa_cfg_v1_e0a3d196b9052321e7008c698e73f936a46f53cfdf612305492866420090cc21","delta":{"denominator":16632,"numerator":61,"percentage_points":"+0.37 pp"},"model_configuration":"gpt-4-turbo-2024-04-09","paired_original":{"denominator":88,"numerator":41,"percent":"46.59%"},"paired_original_rank":192,"protocol":"mean@8","rank_change":-1},{"clean":{"denominator":756,"numerator":355,"percent":"46.96%"},"clean_rank":193,"configuration_id":"gpqa_cfg_v1_e8a07532d90482fd2050087865a6e3e30cf6a9ffb8a09595ff45e6be66586c51","delta":{"denominator":33264,"numerator":227,"percentage_points":"+0.68 pp"},"model_configuration":"Llama-3.1-Tulu-3-70B-DPO","paired_original":{"denominator":1584,"numerator":733,"percent":"46.28%"},"paired_original_rank":193,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":504,"numerator":235,"percent":"46.63%"},"clean_rank":195,"configuration_id":"gpqa_cfg_v1_0aecd21ee0bc485147dfb311f84fdcd13ab5be37a49413da6200f0e3a3f8a9e4","delta":{"denominator":924,"numerator":5,"percentage_points":"+0.54 pp"},"model_configuration":"qwen2.5-32b-instruct","paired_original":{"denominator":792,"numerator":365,"percent":"46.09%"},"paired_original_rank":194,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":1512,"numerator":703,"percent":"46.49%"},"clean_rank":196,"configuration_id":"gpqa_cfg_v1_fb3188e98dcd04325c815261985bb5384ea7240ee9d0ef804de82473aede6bc9","delta":{"denominator":16632,"numerator":299,"percentage_points":"+1.80 pp"},"model_configuration":"DeepSeek-R1-Distill-Qwen-14B","paired_original":{"denominator":132,"numerator":59,"percent":"44.70%"},"paired_original_rank":198,"protocol":"mean@8","rank_change":2},{"clean":{"denominator":28,"numerator":13,"percent":"46.43%"},"clean_rank":197,"configuration_id":"gpqa_cfg_v1_e9fe429ba7834a25f5cf0d8f2deb496f1db6fe7b7552467a7b880fdd61f18cc6","delta":{"denominator":22176,"numerator":125,"percentage_points":"+0.56 pp"},"model_configuration":"gemini-1.5-pro-001","paired_original":{"denominator":3168,"numerator":1453,"percent":"45.86%"},"paired_original_rank":196,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":432,"numerator":199,"percent":"46.06%"},"clean_rank":198,"configuration_id":"gpqa_cfg_v1_3d7b3a0e3e36b74bdde08e2c386e942a53b522bd7c61c1151e6dc007e6e9f001","delta":{"denominator":9504,"numerator":73,"percentage_points":"+0.77 pp"},"model_configuration":"mistral-small-2501","paired_original":{"denominator":3168,"numerator":1435,"percent":"45.30%"},"paired_original_rank":197,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":189,"numerator":85,"percent":"44.97%"},"clean_rank":199,"configuration_id":"gpqa_cfg_v1_dfcacd48f5945f1701b6202adec1a676dc8eb0e19b95f390dfe26319840ae1d5","delta":{"denominator":4158,"numerator":43,"percentage_points":"+1.03 pp"},"model_configuration":"gemma-3-27b-it","paired_original":{"denominator":66,"numerator":29,"percent":"43.94%"},"paired_original_rank":200,"protocol":"mean@1","rank_change":1},{"clean":{"denominator":3024,"numerator":1357,"percent":"44.87%"},"clean_rank":200,"configuration_id":"gpqa_cfg_v1_536a3d475a86a0bde0d95ab2fea86e8ab597b125c21175652a1b26a307d66765","delta":{"denominator":33264,"numerator":227,"percentage_points":"+0.68 pp"},"model_configuration":"Llama-3.1-70B-Instruct","paired_original":{"denominator":396,"numerator":175,"percent":"44.19%"},"paired_original_rank":199,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":3024,"numerator":1339,"percent":"44.28%"},"clean_rank":201,"configuration_id":"gpqa_cfg_v1_b8507ecd7db462de6e723c0dd6cefa055697adc5afe4eaa772a1d3f43c291371","delta":{"denominator":33264,"numerator":281,"percentage_points":"+0.84 pp"},"model_configuration":"WizardLM-2-8x22B","paired_original":{"denominator":99,"numerator":43,"percent":"43.43%"},"paired_original_rank":201,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":1313,"percent":"43.42%"},"clean_rank":202,"configuration_id":"gpqa_cfg_v1_bd2829d168e72420d4610fdbbbcdd955033646500d77a719befb520b54bcc796","delta":{"denominator":189,"numerator":2,"percentage_points":"+1.06 pp"},"model_configuration":"gpt-4-1106-preview","paired_original":{"denominator":144,"numerator":61,"percent":"42.36%"},"paired_original_rank":202,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":189,"numerator":82,"percent":"43.39%"},"clean_rank":203,"configuration_id":"gpqa_cfg_v1_8c7fa3c8345975553ad175f719559715fdda562d71eb9822dec2cf33e6dfeb1e","delta":{"denominator":66528,"numerator":745,"percentage_points":"+1.12 pp"},"model_configuration":"gpt-4-0125-preview","paired_original":{"denominator":3168,"numerator":1339,"percent":"42.27%"},"paired_original_rank":203,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":189,"numerator":79,"percent":"41.80%"},"clean_rank":204,"configuration_id":"gpqa_cfg_v1_a8f7b12d9b5ebb9fdec3e12f1320eef478aff93bb92584ed2185df0ebaf28be5","delta":{"denominator":16632,"numerator":1,"percentage_points":"+0.01 pp"},"model_configuration":"qwen-turbo-2024-11-01","paired_original":{"denominator":792,"numerator":331,"percent":"41.79%"},"paired_original_rank":204,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":3024,"numerator":1259,"percent":"41.63%"},"clean_rank":205,"configuration_id":"gpqa_cfg_v1_b469234b3adc87ce38b386bcaf40c26f1aa236300b162eae3f193b325d11ed95","delta":{"denominator":33264,"numerator":199,"percentage_points":"+0.60 pp"},"model_configuration":"Llama-3.2-90B-Vision-Instruct","paired_original":{"denominator":792,"numerator":325,"percent":"41.04%"},"paired_original_rank":205,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":1008,"numerator":419,"percent":"41.57%"},"clean_rank":206,"configuration_id":"gpqa_cfg_v1_58e2ae82542f35fedd02586ab2a506dc9dd9a92ad9f1913f2bcb0393aeaf603b","delta":{"denominator":22176,"numerator":223,"percentage_points":"+1.01 pp"},"model_configuration":"Meta-Llama-3-70B-Instruct","paired_original":{"denominator":3168,"numerator":1285,"percent":"40.56%"},"paired_original_rank":208,"protocol":"mean@16","rank_change":2},{"clean":{"denominator":3024,"numerator":1249,"percent":"41.30%"},"clean_rank":207,"configuration_id":"gpqa_cfg_v1_56d66d76f7fd4532bd5045af003c34d1f07f98b76657161a697fa41101524e25","delta":{"denominator":8316,"numerator":59,"percentage_points":"+0.71 pp"},"model_configuration":"claude-3-sonnet-20240229","paired_original":{"denominator":1584,"numerator":643,"percent":"40.59%"},"paired_original_rank":207,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":1237,"percent":"40.91%"},"clean_rank":208,"configuration_id":"gpqa_cfg_v1_6970bd17ac70725a425ec6d86eb63a2d2c878ee8acefa59ba35a8efaa4180714","delta":{"denominator":33264,"numerator":41,"percentage_points":"+0.12 pp"},"model_configuration":"qwen2-72b-instruct","paired_original":{"denominator":792,"numerator":323,"percent":"40.78%"},"paired_original_rank":206,"protocol":"mean@16","rank_change":-2},{"clean":{"denominator":1008,"numerator":407,"percent":"40.38%"},"clean_rank":209,"configuration_id":"gpqa_cfg_v1_1d957091393fa4b70f270be10a21999490fcaeae1de9aa2a1d5f9a5196c80ea0","delta":{"denominator":22176,"numerator":1,"percentage_points":"0.00 pp"},"model_configuration":"gemini-1.5-flash-001","paired_original":{"denominator":3168,"numerator":1279,"percent":"40.37%"},"paired_original_rank":209,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":378,"numerator":149,"percent":"39.42%"},"clean_rank":210,"configuration_id":"gpqa_cfg_v1_4c7af7da68de4f30e4d7cddcc9db1fbbf77833670cae85770afbb589903f3414","delta":{"denominator":16632,"numerator":109,"percentage_points":"+0.66 pp"},"model_configuration":"mistral-large-2402","paired_original":{"denominator":792,"numerator":307,"percent":"38.76%"},"paired_original_rank":210,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":1512,"numerator":589,"percent":"38.96%"},"clean_rank":211,"configuration_id":"gpqa_cfg_v1_28ad4b722e7c76693ee321fb599eaa8bcce2126afeffed3cb5a53f356a229340","delta":{"denominator":16632,"numerator":137,"percentage_points":"+0.82 pp"},"model_configuration":"claude-3-5-haiku-20241022","paired_original":{"denominator":396,"numerator":151,"percent":"38.13%"},"paired_original_rank":211,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":1512,"numerator":587,"percent":"38.82%"},"clean_rank":212,"configuration_id":"gpqa_cfg_v1_d212f3b7012127430e247ff410264d8f3ad2193551b8da2fd4265bff806d73fa","delta":{"denominator":66528,"numerator":733,"percentage_points":"+1.10 pp"},"model_configuration":"gpt-4o-mini-2024-07-18","paired_original":{"denominator":3168,"numerator":1195,"percent":"37.72%"},"paired_original_rank":212,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":1157,"percent":"38.26%"},"clean_rank":213,"configuration_id":"gpqa_cfg_v1_322dfa16da9ed38170bad2a7f7bdc94967cda750fa07cddee0c558f1bd8d506e","delta":{"denominator":66528,"numerator":527,"percentage_points":"+0.79 pp"},"model_configuration":"Hermes-2-Theta-Llama-3-70B","paired_original":{"denominator":3168,"numerator":1187,"percent":"37.47%"},"paired_original_rank":213,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":756,"numerator":275,"percent":"36.38%"},"clean_rank":214,"configuration_id":"gpqa_cfg_v1_8dc9a927af2441ac3a54b771539215cedc19d0bdaf6864c0faff20a39f9e062f","delta":{"denominator":33264,"numerator":25,"percentage_points":"+0.08 pp"},"model_configuration":"claude-3-haiku-20240307","paired_original":{"denominator":1584,"numerator":575,"percent":"36.30%"},"paired_original_rank":215,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":378,"numerator":137,"percent":"36.24%"},"clean_rank":215,"configuration_id":"gpqa_cfg_v1_3706e715c29d8cc087e5eea8fcdcfc37f9f912420fe1550b3d630f10b709d0cc","delta":{"denominator":16632,"numerator":-41,"percentage_points":"-0.25 pp"},"model_configuration":"gemma-2-27b-it","paired_original":{"denominator":792,"numerator":289,"percent":"36.49%"},"paired_original_rank":214,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":1512,"numerator":545,"percent":"36.04%"},"clean_rank":216,"configuration_id":"gpqa_cfg_v1_58d9552106e8f5dfbcb2f79f29d67a6ee305a8dc9f8492f63ddf624493d7530a","delta":{"denominator":4158,"numerator":13,"percentage_points":"+0.31 pp"},"model_configuration":"gpt-4-0314","paired_original":{"denominator":792,"numerator":283,"percent":"35.73%"},"paired_original_rank":216,"protocol":"mean@8","rank_change":0},{"clean":{"denominator":1512,"numerator":523,"percent":"34.59%"},"clean_rank":217,"configuration_id":"gpqa_cfg_v1_c030a6e64a8b74f6e73831a3aae768dc618173ab7994e9387b5f3a60cd882de0","delta":{"denominator":2079,"numerator":13,"percentage_points":"+0.63 pp"},"model_configuration":"gemini-1.0-pro-001","paired_original":{"denominator":792,"numerator":269,"percent":"33.96%"},"paired_original_rank":219,"protocol":"mean@16","rank_change":2},{"clean":{"denominator":108,"numerator":37,"percent":"34.26%"},"clean_rank":218,"configuration_id":"gpqa_cfg_v1_29e5aee9510b318690e575f5c2a1337b657fda114d776a3b4bbd40440424339b","delta":{"denominator":9504,"numerator":19,"percentage_points":"+0.20 pp"},"model_configuration":"open-mixtral-8x22b","paired_original":{"denominator":3168,"numerator":1079,"percent":"34.06%"},"paired_original_rank":218,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":1512,"numerator":517,"percent":"34.19%"},"clean_rank":219,"configuration_id":"gpqa_cfg_v1_bb7b9a95d5cff8aa62bb04540cc6a3c48427cc5b28e23bde5a188484ba9cda5d","delta":{"denominator":33264,"numerator":-155,"percentage_points":"-0.47 pp"},"model_configuration":"claude-2.0","paired_original":{"denominator":176,"numerator":61,"percent":"34.66%"},"paired_original_rank":217,"protocol":"mean@16","rank_change":-2},{"clean":{"denominator":3024,"numerator":1033,"percent":"34.16%"},"clean_rank":220,"configuration_id":"gpqa_cfg_v1_724ee2f69b618a701e0d80941d264c74cb5c2349b4b958a51c605beca4d829c3","delta":{"denominator":16632,"numerator":43,"percentage_points":"+0.26 pp"},"model_configuration":"Eurus-2-7B-PRIME","paired_original":{"denominator":528,"numerator":179,"percent":"33.90%"},"paired_original_rank":220,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":756,"numerator":253,"percent":"33.47%"},"clean_rank":221,"configuration_id":"gpqa_cfg_v1_f2ec9c6f943036ef8ca6bef0edc2ea301077976070fe65d840676bbb44703a53","delta":{"denominator":16632,"numerator":85,"percentage_points":"+0.51 pp"},"model_configuration":"claude-2.1","paired_original":{"denominator":88,"numerator":29,"percent":"32.95%"},"paired_original_rank":221,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":1512,"numerator":505,"percent":"33.40%"},"clean_rank":222,"configuration_id":"gpqa_cfg_v1_68be4e024b354a452f5a19040775c2d5aecb864abe7845e7832d52a990885029","delta":{"denominator":33264,"numerator":169,"percentage_points":"+0.51 pp"},"model_configuration":"dbrx-instruct","paired_original":{"denominator":1584,"numerator":521,"percent":"32.89%"},"paired_original_rank":223,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":1512,"numerator":499,"percent":"33.00%"},"clean_rank":223,"configuration_id":"gpqa_cfg_v1_f37332153c1ff0005df8b2ece84a9097d786a435e9046ba52befebedb2adfd3e","delta":{"denominator":2079,"numerator":1,"percentage_points":"+0.05 pp"},"model_configuration":"gemini-1.5-flash-8b-001","paired_original":{"denominator":88,"numerator":29,"percent":"32.95%"},"paired_original_rank":221,"protocol":"mean@8","rank_change":-2},{"clean":{"denominator":1008,"numerator":325,"percent":"32.24%"},"clean_rank":224,"configuration_id":"gpqa_cfg_v1_cf9f9998388c8eddb101a1c6d6fdf01748173bb032842e230ac263e40899daa5","delta":{"denominator":22176,"numerator":59,"percentage_points":"+0.27 pp"},"model_configuration":"Yi-1.5-34B-Chat","paired_original":{"denominator":3168,"numerator":1013,"percent":"31.98%"},"paired_original_rank":224,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":947,"percent":"31.32%"},"clean_rank":225,"configuration_id":"gpqa_cfg_v1_7671edcdc19c9d32865c60525cb98415cb4da59e1b67ca91aa8f1c27e9a823bf","delta":{"denominator":66528,"numerator":443,"percentage_points":"+0.67 pp"},"model_configuration":"gpt-4-0613","paired_original":{"denominator":3168,"numerator":971,"percent":"30.65%"},"paired_original_rank":226,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":16,"numerator":5,"percent":"31.25%"},"clean_rank":226,"configuration_id":"gpqa_cfg_v1_18fba063eae1be42dd8389b74872396a60463f5d291bf70c9532cbf3f5cd9e92","delta":{"denominator":1056,"numerator":7,"percentage_points":"+0.66 pp"},"model_configuration":"Mixtral-8x7B-Instruct-v0.1","paired_original":{"denominator":1056,"numerator":323,"percent":"30.59%"},"paired_original_rank":227,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":216,"numerator":67,"percent":"31.02%"},"clean_rank":227,"configuration_id":"gpqa_cfg_v1_e20f7fe639d9555d87a238b746aca2117c1f00aca9069bf5ace08476f215d92f","delta":{"denominator":4752,"numerator":13,"percentage_points":"+0.27 pp"},"model_configuration":"qwen1.5-32b-chat","paired_original":{"denominator":1584,"numerator":487,"percent":"30.74%"},"paired_original_rank":225,"protocol":"mean@16","rank_change":-2},{"clean":{"denominator":756,"numerator":229,"percent":"30.29%"},"clean_rank":228,"configuration_id":"gpqa_cfg_v1_9ea8f6cc0477f44d90b6a00049e5e0d5848d4514109e686a2ebe2c8c74ade197","delta":{"denominator":66528,"numerator":307,"percentage_points":"+0.46 pp"},"model_configuration":"open-mixtral-8x7b","paired_original":{"denominator":352,"numerator":105,"percent":"29.83%"},"paired_original_rank":229,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":3024,"numerator":907,"percent":"29.99%"},"clean_rank":229,"configuration_id":"gpqa_cfg_v1_39ebb391cd30cb3a52b963dd160804c308e182b68c53d4420593c1241547bbb7","delta":{"denominator":66528,"numerator":67,"percentage_points":"+0.10 pp"},"model_configuration":"open-mistral-nemo-2407","paired_original":{"denominator":3168,"numerator":947,"percent":"29.89%"},"paired_original_rank":228,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":756,"numerator":223,"percent":"29.50%"},"clean_rank":230,"configuration_id":"gpqa_cfg_v1_bcfd3a2eaf724a8a23244b602b5694e0a00e463c8d82eb1c2c4ab4333abe3a1e","delta":{"denominator":6048,"numerator":41,"percentage_points":"+0.68 pp"},"model_configuration":"qwen1.5-72b-chat","paired_original":{"denominator":288,"numerator":83,"percent":"28.82%"},"paired_original_rank":230,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":1008,"numerator":281,"percent":"27.88%"},"clean_rank":231,"configuration_id":"gpqa_cfg_v1_08bd8befe0cade55c7342c9f96b69d027677412b414b670247ac45b3f919f540","delta":{"denominator":11088,"numerator":-17,"percentage_points":"-0.15 pp"},"model_configuration":"gpt-3.5-turbo-1106","paired_original":{"denominator":132,"numerator":37,"percent":"28.03%"},"paired_original_rank":231,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":504,"numerator":139,"percent":"27.58%"},"clean_rank":232,"configuration_id":"gpqa_cfg_v1_79ea3c485cc923af3af9258fe5fa39ad7844f3c46260f2a10709c1ff9e1b57e6","delta":{"denominator":11088,"numerator":-1,"percentage_points":"-0.01 pp"},"model_configuration":"Phi-3-medium-128k-instruct","paired_original":{"denominator":1584,"numerator":437,"percent":"27.59%"},"paired_original_rank":232,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":504,"numerator":139,"percent":"27.58%"},"clean_rank":232,"configuration_id":"gpqa_cfg_v1_c486088d5d896e439d15de33eede35061d90e78929121ccb49dcd5c09ad7b1cd","delta":{"denominator":11088,"numerator":13,"percentage_points":"+0.12 pp"},"model_configuration":"gemma-2-9b-it","paired_original":{"denominator":528,"numerator":145,"percent":"27.46%"},"paired_original_rank":233,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":216,"numerator":59,"percent":"27.31%"},"clean_rank":234,"configuration_id":"gpqa_cfg_v1_4ff2a56599222d2a455b2c7961cc7f544af144e423b64cfd6a02e8ed9c88e446","delta":{"denominator":9504,"numerator":13,"percentage_points":"+0.14 pp"},"model_configuration":"gpt-3.5-turbo-0125","paired_original":{"denominator":1056,"numerator":287,"percent":"27.18%"},"paired_original_rank":234,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":48,"numerator":13,"percent":"27.08%"},"clean_rank":235,"configuration_id":"gpqa_cfg_v1_48c8d1776d8586ba5fce19ed21b2148cc8f42ab5c360ce1f53012f8973abf633","delta":{"denominator":132,"numerator":1,"percentage_points":"+0.76 pp"},"model_configuration":"Llama-2-70b-chat-hf","paired_original":{"denominator":528,"numerator":139,"percent":"26.33%"},"paired_original_rank":236,"protocol":"mean@16","rank_change":1},{"clean":{"denominator":189,"numerator":50,"percent":"26.46%"},"clean_rank":236,"configuration_id":"gpqa_cfg_v1_08f66b9ca9530d2a229621c663831d6ac6cfa888b11230239eb991363750e754","delta":{"denominator":16632,"numerator":-115,"percentage_points":"-0.69 pp"},"model_configuration":"ministral-8b-2410","paired_original":{"denominator":792,"numerator":215,"percent":"27.15%"},"paired_original_rank":235,"protocol":"mean@16","rank_change":-1},{"clean":{"denominator":3024,"numerator":797,"percent":"26.36%"},"clean_rank":237,"configuration_id":"gpqa_cfg_v1_3166c2961eb47a41245ec120bbd6aa00af424fb8d90af97fb2168f58e13fb285","delta":{"denominator":16632,"numerator":47,"percentage_points":"+0.28 pp"},"model_configuration":"Meta-Llama-3-8B-Instruct","paired_original":{"denominator":1584,"numerator":413,"percent":"26.07%"},"paired_original_rank":237,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":793,"percent":"26.22%"},"clean_rank":238,"configuration_id":"gpqa_cfg_v1_b6ca95c2ac417673b46a71c78ad3dd16d60626c176c639d56b530e84b8805233","delta":{"denominator":8316,"numerator":23,"percentage_points":"+0.28 pp"},"model_configuration":"Llama-3.1-8B-Instruct","paired_original":{"denominator":528,"numerator":137,"percent":"25.95%"},"paired_original_rank":238,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":126,"numerator":31,"percent":"24.60%"},"clean_rank":239,"configuration_id":"gpqa_cfg_v1_139cbaca23c931c65e3d28182449928faf23acb7ce5481a3b9bb7577edf379de","delta":{"denominator":154,"numerator":-1,"percentage_points":"-0.65 pp"},"model_configuration":"ministral-3b-2410","paired_original":{"denominator":99,"numerator":25,"percent":"25.25%"},"paired_original_rank":239,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":216,"numerator":53,"percent":"24.54%"},"clean_rank":240,"configuration_id":"gpqa_cfg_v1_937900a08828620bb405325f43551909f02e0fa51fe3dcfd74cfe0f34f2404f2","delta":{"denominator":1188,"numerator":-1,"percentage_points":"-0.08 pp"},"model_configuration":"deepseek-llm-67b-chat","paired_original":{"denominator":264,"numerator":65,"percent":"24.62%"},"paired_original_rank":240,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":3024,"numerator":467,"percent":"15.44%"},"clean_rank":241,"configuration_id":"gpqa_cfg_v1_7f2ac119b9c9f420c06ad2e6f93c6e2e82828c27cec46dd73558c7faf28e5573","delta":{"denominator":66528,"numerator":173,"percentage_points":"+0.26 pp"},"model_configuration":"Mistral-7B-Instruct-v0.3","paired_original":{"denominator":3168,"numerator":481,"percent":"15.18%"},"paired_original_rank":241,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":756,"numerator":115,"percent":"15.21%"},"clean_rank":242,"configuration_id":"gpqa_cfg_v1_5e3089906c8c2437abc826f155bc422a429bb1076e8eaaa770f17c305e3498ab","delta":{"denominator":66528,"numerator":313,"percentage_points":"+0.47 pp"},"model_configuration":"Yi-34B-Chat","paired_original":{"denominator":3168,"numerator":467,"percent":"14.74%"},"paired_original_rank":242,"protocol":"mean@16","rank_change":0},{"clean":{"denominator":84,"numerator":11,"percent":"13.10%"},"clean_rank":243,"configuration_id":"gpqa_cfg_v1_00e9ae21c6c0f764c862027370ab0a348eb54d424a8aff0cb6a305a38618706b","delta":{"denominator":22176,"numerator":-29,"percentage_points":"-0.13 pp"},"model_configuration":"open-mistral-7b","paired_original":{"denominator":3168,"numerator":419,"percent":"13.23%"},"paired_original_rank":243,"protocol":"mean@16","rank_change":0}],"schema_version":"gpqa-clean-results-score-changes-v2","summary":{"all_accepted_configurations":{"configuration_count":243,"decrease_count":8,"definition":"Every canonical accepted direct, compatibility, derived-uniform, targeted-recovery, and independent-native configuration in the combined ranking.","first_quartile_delta":{"denominator":2772,"numerator":29,"percentage_points":"+1.05 pp"},"increase_count":235,"label":"All accepted configurations","median_delta":{"denominator":4158,"numerator":101,"percentage_points":"+2.43 pp"},"third_quartile_delta":{"denominator":2079,"numerator":62,"percentage_points":"+2.98 pp"},"thresholds":[{"clean_count":45,"crossed_down_count":0,"crossed_up_count":20,"paired_original_count":25,"threshold_percent":90},{"clean_count":14,"crossed_down_count":0,"crossed_up_count":14,"paired_original_count":0,"threshold_percent":95}],"unchanged_count":0},"direct_complete_epoch_artifacts":{"configuration_count":224,"decrease_count":8,"definition":"Only the 224 strict or compatibility-complete direct Epoch public artifacts; excludes derived, targeted-recovery, and independent-native configurations.","first_quartile_delta":{"denominator":88704,"numerator":785,"percentage_points":"+0.88 pp"},"increase_count":216,"label":"Direct complete Epoch artifacts","median_delta":{"denominator":8316,"numerator":193,"percentage_points":"+2.32 pp"},"third_quartile_delta":{"denominator":33264,"numerator":977,"percentage_points":"+2.94 pp"},"thresholds":[{"clean_count":31,"crossed_down_count":0,"crossed_up_count":18,"paired_original_count":13,"threshold_percent":90},{"clean_count":6,"crossed_down_count":0,"crossed_up_count":6,"paired_original_count":0,"threshold_percent":95}],"unchanged_count":0}},"summary_contract":{"quantile_input":"Exact rational paired Clean-minus-Original deltas within the named cohort.","quantile_method":"exact linear interpolation at (n-1)q","threshold_method":"Counts use exact rational scores at or above each stated percentage threshold."}}
