Mathematical Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
17.86HMMT Score
6
Apr 20, 2026
43.34Score
6
Apr 20, 2026
18.03Score
6
Apr 20, 2026
48.4True Accuracy
6
Apr 20, 2026
79Accuracy (MATH500)
6
Apr 20, 2026
90Accuracy
6
Apr 20, 2026
3.5∆GSM8K (pp)
6
Apr 20, 2026
54.6Average Score
6
Apr 20, 2026
92.2MATH Accuracy
6
Apr 16, 2026
8Pass@1
6
Apr 15, 2026
36.7Drop in Utility
6
Apr 14, 2026
-7.8Drop in Utility
6
Apr 14, 2026
3Speed Up Ratio
6
Apr 14, 2026
80.51Score
6
Apr 13, 2026
84Pass@8 Rate
6
Apr 10, 2026
65.9Accuracy
6
Apr 10, 2026
33.33Pass@1
6
Apr 10, 2026
73.6Pass@1 Rate
6
Apr 10, 2026
76Accuracy
6
Apr 10, 2026
91Accuracy
6
Apr 10, 2026
23.7Throughput (TPS)
6
Apr 9, 2026
39maj@8
6
Apr 7, 2026
68.4maj@8
6
Apr 7, 2026
43.3maj@8
6
Apr 7, 2026
98.8Accuracy
6
Apr 6, 2026