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

Scientific Problem Solving on PhysicsQA

53.22Accuracy

Reason, Reward, Refine

17.54826.80936.0745.331Jul 6, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.07
53.22
2026.07
49.62
2026.07
46.73
2026.07
41.59
2026.07
41.49
2026.07
39.19
2026.07
39.02
2026.07
36.91
2026.07
33.35
2026.07
31.21
2026.07
30.64
2026.07
28.64
2026.07
27.84
2026.07
27.67
2026.07
25.95
2026.07
24.32
2026.07
23.51
2026.07
23.24
2026.07
21.35
2026.07
18.92