List your company

Jailbreak Resistance on GCG

96.98Refusal RateDR-IRL
54.91265.833576.75587.6765Mar 23, 2025

Evaluation results

MethodLinks
DR-IRLModel=LLaMA-3.1 8B2025.0396.98
STAIRModel=LLaMA-3.1 8B2025.0391.28
BaseModel=LLaMA-3.1 8B2025.0356.53