List your company

Mathematical Reasoning on MATH500 (Acc, CI, Token Usage, Flips)

76.3AccuracySelective active verify
58.30862.97967.6572.321Jun 18, 2026

Evaluation results

MethodLinks
Selective active verifyPolicy strategy=Selective active verify2026.0676.372.848.22,7917,1041
Long baseGeneration limit=8,1922026.067672.2005,1240
Always active verifyPolicy strategy=Always active verify2026.0675.571.91003,8128,1252.2
Selective continuePolicy strategy=Selective continue2026.0673.669.948.22,7517,0641.7
Always continuePolicy strategy=Always continue2026.067268.41003,6948,0073.6
BaseGeneration limit=4,0962026.065954.8004,3130