Mathematical Problem Solving
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
87.78Pass@12
5
May 22, 2026
98.4Pass@12
5
May 22, 2026
97.8Pass@1
5
May 12, 2026
95.56Pass@1
5
May 12, 2026
5Correct Count
5
Apr 21, 2026
32.5Pass@1 Accuracy
5
Mar 4, 2026
96.4Pass@128
5
Feb 26, 2026
93.3Pass@1
5
Feb 26, 2026
14.1Pass@1
5
Feb 26, 2026
75Maj@k
5
Feb 26, 2026
0.308maj@k
5
Feb 26, 2026
84.4Score
4
Jul 2, 2026
96.7Score
4
Jul 2, 2026
95.2Score
4
Jul 2, 2026
57.41Vanilla Accuracy
4
Jun 23, 2026
90.6Mean Score
4
Jun 12, 2026
36.89Throughput (Tokens/s)
4
Jun 4, 2026
69.7Accuracy
4
May 11, 2026
1.22Training Cost (USD)
4
Feb 26, 2026
90pass@50
4
Feb 26, 2026
22.71Avg@16 Acc
4
Feb 26, 2026
26.67Average@16 Accuracy
4
Feb 26, 2026
96.36GSM8K
4
Feb 26, 2026
80Score
4
Feb 26, 2026
83.3Points
4
Jun 23, 2026