List your company

Mathematical Problem Solving on AIME 2025 (Pass@k Series)

12.9Pass@1CaSP
5.3087.2799.2511.221Oct 16, 2025

Evaluation results

MethodLinks
CaSPBackbone=Qwen2.5-Math-7B2025.1012.919.225.430.935.840.444.749.555.662.568.9
GRPOBackbone=Qwen2.5-Math-7B2025.101217.323.429.534.739.243.547.953.158.362.3
Base ModelBackbone=Qwen2.5-Math-7B2025.105.69.514.720.82732.938.644.350.557.965.5