List your company

Mathematical Reasoning on AIME24 (mean@8)

39.79Mean@8 ScoreACPO
21.912426.553731.19535.8363Jul 3, 2026

Evaluation results

MethodLinks
ACPOModel=DeepSeek-R1-Distill-Qwen-7B2026.0739.79
SAPOModel=DeepSeek-R1-Distill-Qwen-7B2026.0738.54
ACPOModel=Qwen2.5-Math-7B2026.0736.56
GTPOModel=DeepSeek-R1-Distill-Qwen-7B2026.0736.37
GTPOModel=Qwen2.5-Math-7B2026.0733.97
SAPOModel=Qwen2.5-Math-7B2026.0732.29
80/20Model=DeepSeek-R1-Distill-Qwen-7B2026.0732.16
ACPOModel=Qwen3-8B-Base2026.0730.31
80/20Model=Qwen2.5-Math-7B2026.0729.25
SAPOModel=Qwen3-8B-Base2026.0729.06
DAPOModel=DeepSeek-R1-Distill-Qwen-7B2026.0729.06
GRPOModel=DeepSeek-R1-Distill-Qwen-7B2026.0728.43
DAPOModel=Qwen2.5-Math-7B2026.0727.7
GTPOModel=Qwen3-8B-Base2026.0727.21
GRPOModel=Qwen2.5-Math-7B2026.0726.25
80/20Model=Qwen3-8B-Base2026.0726.02
DAPOModel=Qwen3-8B-Base2026.0724.89
GRPOModel=Qwen3-8B-Base2026.0722.6