List your company
ResearchFormerly WizWand

Refusal Evaluation on Past Tense

51ASRQwen2.5-7B-Instruct
-2.0411.7325.539.27Sep 30, 2025
Updated 6mo ago

Evaluation results

MethodLinks
Qwen2.5-7B-InstructBase Model=Qwen2.5-7B-Instruct2025.0951
DPOBase Model=Qwen2.5-7B-Instruct2025.0949
SFT (5/95)Base Model=Qwen2.5-7B-Instruct2025.0947
CBBase Model=Qwen2.5-7B-Instruct2025.0947
RepEBase Model=Qwen2.5-7B-Instruct2025.0946
Llama-3.1-8B-InstructBase Model=Llama-3.1-8B-Instruct2025.0942
RepEBase Model=Llama-3.1-8B-Instruct2025.0941
Head AblationBase Model=Qwen2.5-7B-Instruct2025.0941
DPOBase Model=Llama-3.1-8B-Instruct2025.0938
Gemma-2-9B-itBase Model=Gemma-2-9B-it2025.0938
Only ScalingBase Model=Qwen2.5-7B-Instruct2025.0937
DPOBase Model=Gemma-2-9B-it2025.0937
CBBase Model=Gemma-2-9B-it2025.0936
Head AblationBase Model=Gemma-2-9B-it2025.0934
RepEBase Model=Gemma-2-9B-it2025.0934
RepBendBase Model=Qwen2.5-7B-Instruct2025.0930
Head AblationBase Model=Llama-3.1-8B-Instruct2025.0929
OLMo-2-1124-7B-InstructBase Model=OLMo-2-1124-7B-Instruct2025.0928
RepBendBase Model=Gemma-2-9B-it2025.0927
Only ScalingBase Model=Gemma-2-9B-it2025.0926
DPOBase Model=OLMo-2-1124-7B-Instruct2025.0925
RepBendBase Model=OLMo-2-1124-7B-Instruct2025.0923
Head AblationBase Model=OLMo-2-1124-7B-Instruct2025.0922
RepEBase Model=OLMo-2-1124-7B-Instruct2025.0922
SFT (5/95)Base Model=Llama-3.1-8B-Instruct2025.0921
SFT (5/95)Base Model=OLMo-2-1124-7B-Instruct2025.0921
CBBase Model=OLMo-2-1124-7B-Instruct2025.0920
ASGUARDBase Model=Gemma-2-9B-it2025.0919
Only ScalingBase Model=OLMo-2-1124-7B-Instruct2025.0917
Only ScalingBase Model=Llama-3.1-8B-Instruct2025.0913
RepBendBase Model=Llama-3.1-8B-Instruct2025.0911
ASGUARDBase Model=OLMo-2-1124-7B-Instruct2025.099
ASGUARDBase Model=Llama-3.1-8B-Instruct2025.098
ASGUARDBase Model=Qwen2.5-7B-Instruct2025.098
SFT (30/70)Base Model=OLMo-2-1124-7B-Instruct2025.098
SFT (30/70)Base Model=Llama-3.1-8B-Instruct2025.093
CBBase Model=Llama-3.1-8B-Instruct2025.090
SFT (30/70)Base Model=Qwen2.5-7B-Instruct2025.090
SFT (5/95)Base Model=Gemma-2-9B-it2025.090
SFT (30/70)Base Model=Gemma-2-9B-it2025.090