SciKnowEval
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
SciKnowEval
82.38Chemistry Accuracy
56
SciKnowEval Chemistry
94.3mean@16
27
SciKnowEval Physics
63.39Mean@16
24
SciKnowEval Material
71.48Mean@16
24
SciKnowEval Biology
37.08Mean@16
24
SciKnowEval L3
81Chemistry Score
18
SciKnowEval
68.5Accuracy (Biology)
15
SciKnowEval
52.1Accuracy
12
SciKnowEval Edu. MMLU-Pro (test)
21.1Average Calibration Score
9
SciKnowEval (test)
0.226BIO Score
9
SciKnowEval Phys-Chem (test)
58.8Accuracy (Physics)
8
SciKnowEval Chem-Mat (test)
71.3Chemistry Accuracy
8
SciKnowEval Mat-Phys (test)
70.5Accuracy (Mat. Sci.)
8
SciKnowEval reasoning 8
79.61Chemistry Score
6
SciKnowEval Hard
42.4Pass@1
5
SciKnowEval Hardest
21Pass@1
5
SciKnowEval (test)
65.38Biology Accuracy
4
SciKnowEval Physics (val)
93.8Best@16
3
SciKnowEval Material (val)
87.9Best@16
3
SciKnowEval Biology (val)
80.1Best@16
3
SciKnowEval Chemistry L3 (test)
80.6Final Test Mean Score@16
3