Multi-discipline Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
36.1Accuracy
83
Jun 26, 2026
81.8Accuracy
38
Jun 1, 2026
52.2Accuracy
21
Mar 16, 2026
51.8Pass@1
19
Feb 26, 2026
64.4pass@1
19
Feb 26, 2026
47.1Accuracy@8
16
Mar 4, 2026
74.3MMMU Score
14
Feb 26, 2026
24.6Accuracy
8
Feb 26, 2026
17.68Performance Gain Sum
6
Feb 26, 2026