Math
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
67.2Original Accuracy
6
Jun 19, 2026
73.4Accuracy
6
Jun 1, 2026
70.5Accuracy
6
Feb 26, 2026
73.63Accuracy
6
Feb 26, 2026
52.67Accuracy
6
Feb 26, 2026
99Score
5
Jul 2, 2026
97.5Score
5
Jul 2, 2026
69.9Comp. Acc.
5
Mar 18, 2026
14Mean@128
5
Feb 26, 2026
87.66Score (%)
5
Feb 26, 2026
29.21Average Performance
4
May 12, 2026
30.56Average Performance
4
May 12, 2026
32.2Score (%)
4
May 8, 2026
71.6Score
4
May 8, 2026
89.1Score
4
May 8, 2026
94.62GSM8K
4
Feb 26, 2026
68.76Surrogate Accuracy (Initial)
4
Feb 26, 2026
15.7Accuracy
3
May 19, 2026
53.33Pass@32
3
Apr 15, 2026
70Exact Match (EM)
3
Apr 15, 2026
36Pass@1
3
Feb 26, 2026
97.1Accuracy (%)
2
Mar 27, 2026
39Accuracy
2
Feb 26, 2026
81.2Score
2
Feb 26, 2026
0.003Baseline Score
1
Apr 10, 2026