Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
45.41Ours Accuracy
4
Feb 26, 2026
35.2Score
3
Jul 7, 2026
96.6Score
3
Jul 7, 2026
0.57Accuracy (Reasoning)
3
Jun 11, 2026
58.25Accuracy (CEVAL Reasoning)
3
Jun 11, 2026
1.09Speedup (Batch 1)
3
Jun 2, 2026
2.21Speedup (batch-1)
3
Jun 2, 2026
3.08Speedup (batch=1)
3
Jun 2, 2026
3.4Speedup (batch=1)
3
Jun 2, 2026
43.6Accuracy
3
May 25, 2026
72.9Accuracy
3
May 25, 2026
18.15Accuracy
3
May 19, 2026
35.13Accuracy
3
May 19, 2026
100BBH: Boolean Expressions (Mean)
3
May 15, 2026
93.3Copa (HE)
3
May 13, 2026
30AIME 2024 Accuracy
3
May 11, 2026
48.4Accuracy
3
May 8, 2026
78.32Mean Accuracy
3
May 8, 2026
26.15Acc (Normalized)
3
Apr 23, 2026
36.44Normalized Accuracy
3
Apr 23, 2026
58.5Avg@8 Score
3
Apr 23, 2026
71.72Avg@32
3
Apr 23, 2026
41.28Accuracy
3
Apr 21, 2026
69.65Accuracy (zero-shot)
3
Apr 21, 2026
41.9BBH Accuracy
3
Apr 13, 2026