Mathematical Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
50.4Accuracy
2
May 21, 2026
2.93Rollout Speedup Factor
2
May 20, 2026
61.4Accuracy
2
May 20, 2026
0.292Score
2
May 15, 2026
0.6Performance Ratio (PR)
2
May 13, 2026
0.18Token Efficiency Ratio
2
May 11, 2026
25.34Pass@1
2
May 11, 2026
2.5Delta Accuracy (pp)
2
May 8, 2026
49.6Avg@16 Accuracy
2
May 4, 2026
94.5Accuracy (Avg@16)
2
May 4, 2026
45.6Avg@16 Accuracy
2
May 4, 2026
60.9Avg@16 Accuracy
2
May 4, 2026
89.8Accuracy
2
May 4, 2026
97.19Accuracy
2
May 4, 2026
37.1Final Test Accuracy
2
May 1, 2026
97.8Raw ASR
2
Apr 28, 2026
62.8Accuracy
2
Apr 28, 2026
18.06Accuracy
2
Apr 27, 2026
2,025,345Avg Token Usage (R0)
2
Apr 24, 2026
72.5NFE
2
Apr 23, 2026
50.5NFE
2
Apr 23, 2026
45.66Avg@64
2
Apr 21, 2026
75.93Avg Score @64
2
Apr 21, 2026
0.203Det.
2
Apr 21, 2026
13.9Det. Accuracy
2
Apr 21, 2026