Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

SciKnowEval

Benchmarks

Task NameDataset NameSOTA ResultTrend
Scientific ReasoningSciKnowEval
Chemistry Accuracy82.38
56
Scientific ReasoningSciKnowEval Chemistry
mean@1694.3
27
Scientific ReasoningSciKnowEval Physics
Mean@1663.39
24
Scientific ReasoningSciKnowEval Material
Mean@1671.48
24
Scientific ReasoningSciKnowEval Biology
Mean@1637.08
24
Scientific ReasoningSciKnowEval L3
Chemistry Score81
18
Scientific Question AnsweringSciKnowEval
Accuracy (Biology)68.5
15
Scientific Knowledge EvaluationSciKnowEval
Accuracy52.1
12
CalibrationSciKnowEval Edu. MMLU-Pro (test)
Average Calibration Score21.1
9
CalibrationSciKnowEval (test)
BIO Score0.226
9
Multi-domain Science Q&ASciKnowEval Phys-Chem (test)
Accuracy (Physics)58.8
8
Multi-domain Science Q&ASciKnowEval Chem-Mat (test)
Chemistry Accuracy71.3
8
Multi-domain Science Q&ASciKnowEval Mat-Phys (test)
Accuracy (Mat. Sci.)70.5
8
Scientific ReasoningSciKnowEval reasoning 8
Chemistry Score79.61
6
Scientific reasoningSciKnowEval Hard
Pass@142.4
5
Scientific reasoningSciKnowEval Hardest
Pass@121
5
Scientific ReasoningSciKnowEval (test)
Biology Accuracy65.38
4
Scientific ReasoningSciKnowEval Physics (val)
Best@1693.8
3
Scientific ReasoningSciKnowEval Material (val)
Best@1687.9
3
Scientific ReasoningSciKnowEval Biology (val)
Best@1680.1
3
Multiple-choice reasoningSciKnowEval Chemistry L3 (test)
Final Test Mean Score@1680.6
3
Showing 21 of 21 rows