List your company

Mathematical Reasoning on AIME 2025 (Avg@32, Pass@8, Dist-N, SemDiv)

65.21Avg@32CraEG
64.991665.048365.10565.1617Jan 30, 2026

Evaluation results

MethodLinks
CraEGModel=Qwen3-4B, Decoding configuration=Temp = 1.0, Top-p = 1.02026.0165.2182.4755.3417.36
Standard SamplingModel=Qwen3-4B, Decoding configuration=Temp = 1.0, Top-p = 1.02026.016581.7952.3716.3