Physical Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
82.5Accuracy
90
May 20, 2026
26.14Success Rate
36
Mar 17, 2026
47.06Success Rate
36
Mar 17, 2026
36.5Att. 1 Score
36
Mar 17, 2026
26.89Att. 1 Score
36
Mar 17, 2026
37.6Attention Score 1
36
Mar 17, 2026
14.92Att. 1 Score
36
Mar 17, 2026
91.3Accuracy
34
Feb 26, 2026
48.6Accuracy
24
May 15, 2026
56.1AVG Score
22
Feb 26, 2026
82.21Accuracy
20
Feb 26, 2026
29.6Accuracy
20
May 14, 2026
97.1S1 Score
13
Apr 28, 2026
79.71PIQA Normalized Performance
12
May 12, 2026
5.3Pass@1 Accuracy
12
Feb 26, 2026
87.9Pass@1 Accuracy
12
Feb 26, 2026
21.7Pass@1 Accuracy
12
Feb 26, 2026
56.2Coherence
11
May 7, 2026
0.386Accuracy
8
Feb 26, 2026
64.7Overall Score
8
Feb 26, 2026
29.59Physics-IQ Score
7
Apr 10, 2026
42.2AUCCESS
7
Feb 26, 2026
86.2AUCCESS
7
Feb 26, 2026
50.86Success Rate
6
Feb 26, 2026
85.49Success Rate (AUCCESS)
6
Feb 26, 2026