List your company

Overrefusal Evaluation on Fortress OR

97.6Helpfulness ScoreRECAP
81.79285.8969094.104Oct 1, 2025

Evaluation results

MethodLinks
RECAPBackbone=DSQwen-14B2025.1097.6
SafeChainBackbone=DSQwen-14B2025.1096.4
SFTBackbone=DSQwen-14B2025.1096
OriginalBackbone=DSQwen-14B2025.1095
DAPOBackbone=DSQwen-14B2025.1095
STARBackbone=DSQwen-14B2025.1093.2
RECAPBackbone=DSLlama-8B2025.1091.8
OriginalBackbone=DSLlama-8B2025.1090
STARBackbone=DSLlama-8B2025.1086
SafeChainBackbone=DSLlama-8B2025.1084.5
DAPOBackbone=DSLlama-8B2025.1082.8
SFTBackbone=DSLlama-8B2025.1082.4