List your company

Tool-integrated Mathematical Reasoning on BeyondAIME (Pass@1, TCN, WTN)

41Pass@1Qwen3-14B + PRUNETIR
30.633.33638.7May 11, 2026

Evaluation results

MethodLinks
Qwen3-14B + PRUNETIRPRUNETIR=true2026.05412.610.1
Qwen3-14BPRUNETIR=false2026.0538.13.110.6
Qwen3-8B + PRUNETIRPRUNETIR=true2026.0535.42.910.8
ReTool-32B + PRUNETIRPRUNETIR=true2026.0532.67.34.3
ReTool-32BPRUNETIR=false2026.0531.46.55.5
Qwen3-8BPRUNETIR=false2026.05316.412.2