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

Physics Reasoning on Public Physics Benchmarks Suite (test)

64.4Average Accuracy

Reason, Reward, Refine

10.923224.806638.6952.5734May 16, 2026May 24, 2026Jun 2, 2026Jun 10, 2026Jun 19, 2026Jun 27, 2026Jul 6, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.07
64.4---
2026.07
61.5---
2026.07
58.5---
2026.07
54.3---
2026.07
53.9---
2026.05
53.4253.4953.7153.05
2026.07
53.3---
2026.05
52.2660.4655.9540.36
2026.07
52.1---
2026.07
51.2---
2026.07
51---
2026.07
50.5---
2026.05
47.3456.5952.3333.09
2026.07
46.1---
2026.05
45.0443.0249.2242.88
2026.07
44.8---
2026.05
44.2751.5550.7730.5
2026.07
44.1---
2026.07
44---
2026.05
43.7453.4950.9426.8
2026.05
41.4147.6738.8637.71
2026.07
41.1---
2026.07
41---
2026.07
40.7---
2026.05
37.9846.937.8229.21
2026.07
37.5---
2026.07
37.5---
2026.05
36.2937.9848.722.18
2026.05
35.1439.5335.7530.13
2026.05
35.0843.0232.6429.57
2026.05
34.7241.4743.5219.17
2026.07
32.5---
2026.05
30.9840.727.4624.77
2026.05
29.3936.8232.1219.22
2026.05
28.0336.8226.9420.33
2026.05
23.7123.2630.0517.81
2026.05
20.634.57.1420.15
2026.05
19.4627.0215.0216.33
2026.05
19.2230.629.8417.19
2026.05
13.5819.777.2513.72
2026.05
12.987.3617.0914.48