Mathematical Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
67.43Math Avg. Pass@4 (Before)
4
May 8, 2026
67.9Pass@1
4
May 7, 2026
3.26Acceptance Length (τ)
4
May 7, 2026
420.7Throughput (tok/s)
4
May 7, 2026
28.33AIME24 Accuracy
4
May 1, 2026
55.9Accuracy
4
May 1, 2026
24.9Final Test Accuracy
4
May 1, 2026
79.8Pass@1 Accuracy
4
Apr 30, 2026
78Accuracy
4
Apr 29, 2026
88.8R0 Coverage
4
Apr 24, 2026
41.5R0 Score
4
Apr 24, 2026
94.4Answer Selection Rate (ASR)
4
Apr 23, 2026
91.1ASR
4
Apr 23, 2026
81.022Accuracy (Easy)
4
Apr 23, 2026
71.764Accuracy (Easy)
4
Apr 23, 2026
42.92Pass@1
4
Apr 21, 2026
55.83Pass@1
4
Apr 21, 2026
37.5Accuracy
4
Apr 16, 2026
91.1MATH Accuracy
4
Apr 16, 2026
70.8Accuracy (Ref)
4
Apr 15, 2026
80.5SVAMP Accuracy (Ref)
4
Apr 15, 2026
91.7Reference Accuracy
4
Apr 15, 2026
32.2Accuracy (Omni-MATH)
4
Apr 15, 2026
43.34Accuracy
4
Apr 14, 2026
54.16Accuracy
4
Apr 14, 2026