List your company
ResearchFormerly WizWand

Refusal Evaluation on OR-Bench

99.6Toxic Refusal RateSFT (30/70)
74.43280.96687.594.034Sep 30, 2025
Updated 6mo ago

Evaluation results

MethodLinks
SFT (30/70)Base Model=OLMo-2-1124-7B-Instruct2025.0999.691.8
SFT (30/70)Base Model=Qwen2.5-7B-Instruct2025.0999.598.5
SFT (5/95)Base Model=Gemma-2-9B-it2025.0999.389
ASGUARDBase Model=Gemma-2-9B-it2025.099988
RepBendBase Model=Gemma-2-9B-it2025.0998.984.7
SFT (30/70)Base Model=Gemma-2-9B-it2025.0998.794.9
ASGUARDBase Model=Qwen2.5-7B-Instruct2025.099870.5
ASGUARDBase Model=OLMo-2-1124-7B-Instruct2025.0997.569.2
Head AblationBase Model=Gemma-2-9B-it2025.0997.273.6
RepEBase Model=Gemma-2-9B-it2025.0997.170.7
Only ScalingBase Model=Llama-3.1-8B-Instruct2025.0996.966.2
CBBase Model=Gemma-2-9B-it2025.0996.971.1
Gemma-2-9B-itBase Model=Gemma-2-9B-it2025.0996.770.5
DPOBase Model=Gemma-2-9B-it2025.0996.666.8
ASGUARDBase Model=Llama-3.1-8B-Instruct2025.0996.466.8
RepBendBase Model=Llama-3.1-8B-Instruct2025.0996.177.9
SFT (5/95)Base Model=OLMo-2-1124-7B-Instruct2025.0994.857.9
Only ScalingBase Model=Qwen2.5-7B-Instruct2025.0994.342.6
SFT (5/95)Base Model=Llama-3.1-8B-Instruct2025.0994.150.8
DPOBase Model=OLMo-2-1124-7B-Instruct2025.0993.248.3
Only ScalingBase Model=OLMo-2-1124-7B-Instruct2025.0992.848.8
OLMo-2-1124-7B-InstructBase Model=OLMo-2-1124-7B-Instruct2025.0992.543.5
Head AblationBase Model=OLMo-2-1124-7B-Instruct2025.0992.545.1
CBBase Model=OLMo-2-1124-7B-Instruct2025.0992.343.1
RepBendBase Model=OLMo-2-1124-7B-Instruct2025.099242.8
SFT (30/70)Base Model=Llama-3.1-8B-Instruct2025.0991.980.3
Only ScalingBase Model=Gemma-2-9B-it2025.0991.972.4
RepEBase Model=OLMo-2-1124-7B-Instruct2025.0991.443.2
SFT (5/95)Base Model=Qwen2.5-7B-Instruct2025.0991.138.8
DPOBase Model=Llama-3.1-8B-Instruct2025.0990.233.2
Llama-3.1-8B-InstructBase Model=Llama-3.1-8B-Instruct2025.0988.528.9
RepEBase Model=Llama-3.1-8B-Instruct2025.0987.929.7
Head AblationBase Model=Llama-3.1-8B-Instruct2025.0986.734.7
CBBase Model=Llama-3.1-8B-Instruct2025.0985.184.8
Head AblationBase Model=Qwen2.5-7B-Instruct2025.0980.114.6
CBBase Model=Qwen2.5-7B-Instruct2025.0979.813.1
Qwen2.5-7B-InstructBase Model=Qwen2.5-7B-Instruct2025.0979.512.9
DPOBase Model=Qwen2.5-7B-Instruct2025.0979.315
RepEBase Model=Qwen2.5-7B-Instruct2025.0979.312.8
RepBendBase Model=Qwen2.5-7B-Instruct2025.0975.412.2