List your company

Safety Evaluation on HarmBench Static Context Evaluation Single-turn

70.5SFRCodeAttack
0.318.52536.7554.975Mar 15, 2026

Evaluation results

MethodLinks
CodeAttackTarget Model=GPT 4o, Access Type=Closed-Source Models2026.0370.5
CodeAttackTarget Model=Llama 3-70B-IT, Access Type=Open-Weight Models2026.0366
CodeAttackTarget Model=Llama 3-8B-IT, Access Type=Open-Weight Models2026.0346
CodeAttackTarget Model=Claude 3.5 Sonnet, Access Type=Closed-Source Models2026.0339.5
PAIRTarget Model=GPT 4o, Access Type=Closed-Source Models2026.0339
PAIRTarget Model=Llama 3-70B-IT, Access Type=Open-Weight Models2026.0336
GCGTarget Model=Llama 3-8B-IT, Access Type=Open-Weight Models2026.0334.5
PAIRTarget Model=Llama 3-8B-IT, Access Type=Open-Weight Models2026.0318.7
GCGTarget Model=Llama 3-70B-IT, Access Type=Open-Weight Models2026.0317
GCGTarget Model=GPT 4o, Access Type=Closed-Source Models2026.0312.5
GCGTarget Model=Claude 3.5 Sonnet, Access Type=Closed-Source Models2026.033
PAIRTarget Model=Claude 3.5 Sonnet, Access Type=Closed-Source Models2026.033