List your company
ResearchFormerly WizWand

Jailbreak Attempt Forgetting on Harm Jailbreak 2

71.4ASRRetainOnlyFT
-2.85616.42235.754.978Apr 7, 2026
Updated 6mo ago

Evaluation results

MethodLinks
RetainOnlyFTModel=Llama-3.2-1B-Instruct2026.0471.4
RetainOnlyFTModel=OLMo-2-1B-Instruct2026.0452.1
Llama-3.2-1B-InstructModel=Llama-3.2-1B-Instruct2026.0451.1
RetainOnlyFTModel=OLMo-2-7B-Instruct2026.0446.2
EraserModel=OLMo-2-1B-Instruct, Hyperparameter (λ)=0.62026.0445.6
UnlearningModel=OLMo-2-7B-Instruct, Hyperparameter (λ)=0.62026.0441.7
EraserModel=OLMo-2-7B-Instruct, Hyperparameter (λ)=0.62026.0426.9
OLMo-2-7B-InstructModel=OLMo-2-7B-Instruct2026.0426.6
DPOModel=OLMo-2-7B-Instruct2026.0426.5
DPOModel=OLMo-2-1B-Instruct2026.0418.9
RetainOnlyFTModel=Llama-3.2-3B-Instruct2026.0418.7
DPOModel=Llama-3.2-1B-Instruct2026.0412.8
OLMo-2-1B-InstructModel=OLMo-2-1B-Instruct2026.0411.1
Llama-3.2-3B-InstructModel=Llama-3.2-3B-Instruct2026.0410.2
UnlearningModel=Llama-3.2-3B-Instruct, Hyperparameter (λ)=0.62026.049.1
SKUModel=OLMo-2-7B-Instruct2026.046.5
SKUModel=Llama-3.2-1B-Instruct2026.046.4
SKUModel=Llama-3.2-3B-Instruct2026.046.4
DPOModel=Llama-3.2-3B-Instruct2026.045.4
SKUModel=OLMo-2-1B-Instruct2026.044
UnlearningModel=Llama-3.2-1B-Instruct, Hyperparameter (λ)=0.62026.040.4
EUModel=Llama-3.2-3B-Instruct, Hyperparameter (λ)=0.62026.040.3
EraserModel=Llama-3.2-1B-Instruct, Hyperparameter (λ)=0.62026.040
EUModel=Llama-3.2-1B-Instruct, Hyperparameter (λ)=0.62026.040
EraserModel=Llama-3.2-3B-Instruct, Hyperparameter (λ)=0.62026.040
UnlearningModel=OLMo-2-1B-Instruct, Hyperparameter (λ)=0.62026.040
EUModel=OLMo-2-1B-Instruct, Hyperparameter (λ)=0.62026.040
EUModel=OLMo-2-7B-Instruct, Hyperparameter (λ)=0.62026.040