List your company

Mathematical Reasoning on GSM8K (pass@K)

97.77pass@KRECAP
90.02292.033594.04596.0565Oct 1, 2025

Evaluation results

MethodLinks
RECAPBackbone=DSQwen-14B2025.1097.77
DAPOBackbone=DSQwen-14B2025.1097.19
SafeChainBackbone=DSQwen-14B2025.1096.44
SFTBackbone=DSQwen-14B2025.1095.9
STARBackbone=DSQwen-14B2025.1095.86
OriginalBackbone=DSQwen-14B2025.1095.2
RECAPBackbone=DSLlama-8B2025.1093.72
DAPOBackbone=DSLlama-8B2025.1093.71
SafeChainBackbone=DSLlama-8B2025.1091.32
SFTBackbone=DSLlama-8B2025.1091.32
STARBackbone=DSLlama-8B2025.1090.74
OriginalBackbone=DSLlama-8B2025.1090.32