List your company

Jailbreak Attack on JailbreakBench PAIR

82Attack Success Rate (ASR)None
-3.2818.864163.14Jan 30, 2024Jun 12, 2024Oct 24, 2024Mar 7, 2025Jul 19, 2025Nov 30, 2025Apr 13, 2026

Evaluation results

MethodLinks
NoneModel=Vicuna2024.0182
Perplexity FilterModel=Vicuna2024.0181
SmoothLLMModel=Vicuna2024.0147
w/o DefenseTarget Model=GPT-4o2026.0430
RephrasingModel=Vicuna2024.0125
RPOModel=Vicuna2024.0120
Smooth LLMTarget Model=GPT-4o2026.0419
Avg. ASRTarget Model=GPT-4o2026.0416
Self ReminderTarget Model=GPT-4o2026.0411
NoneModel=Llama-22024.014
Perplexity FilterModel=Llama-22024.014
RephrasingModel=Llama-22024.014
SmoothLLMModel=Llama-22024.011
Input FilterTarget Model=GPT-4o2026.041
RPOModel=Llama-22024.010