List your company

Mathematical Reasoning on Olympiad (Acc@16, Tokens)

39.8Acc@16Self-Verify
19.83225.01630.235.384Feb 7, 2026

Evaluation results

MethodLinks
Self-VerifyModel=Qwen2.5-7B-Instruct2026.0239.81,168
GenerateModel=Qwen2.5-7B-Instruct2026.0237.84,543
Self-VerifyModel=Qwen2.5-3B-Instruct2026.0232.11,891
GenerateModel=Qwen2.5-3B-Instruct2026.0227.43,299
Self-VerifyModel=Qwen2.5-1.5B-Instruct2026.0222.21,351
GenerateModel=Qwen2.5-1.5B-Instruct2026.0220.61,639