Mathematical Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
59.4Error Rate
1
May 28, 2026
45.8Error Rate
1
May 28, 2026
40.1Error Rate
1
May 28, 2026
96.67Initial Accuracy
1
May 26, 2026
73.67Initial Accuracy
1
May 26, 2026
78.4Initial Accuracy
1
May 26, 2026
94.41GSM8K Flexible Extraction Accuracy
1
May 15, 2026
0Accuracy
1
May 12, 2026
87.7Baseline Accuracy
1
May 12, 2026
19.39Avg@64
1
Apr 21, 2026
93.4Detection Rate
1
Apr 21, 2026
78.8Base Accuracy
1
Apr 16, 2026
72.9Pass@1 Score
1
Mar 20, 2026
35Points
1
Mar 20, 2026
70Monotonicity Rate
1
Mar 20, 2026
73.7Monotonicity Rate
1
Mar 20, 2026
31.75Avg@5 Accuracy (Seeds)
1
Feb 26, 2026
58.34Avg@5 Accuracy (Seeds)
1
Feb 26, 2026
0.849Avg@5 Accuracy (Seeds)
1
Feb 26, 2026
58.34Avg@5 Accuracy (Seeds)
1
Feb 26, 2026
79.21Avg@5 Acc (Seeds)
1
Feb 26, 2026
84.46Avg Accuracy @5 (Seeds)
1
Feb 26, 2026
94.85Accuracy (Avg@5, Seeds)
1
Feb 26, 2026
48.9Win Rate
1
Feb 26, 2026
39.4GSM8K Score
1
Feb 26, 2026