List your company

Jailbreak Attack on AdvBench Text

98Total Success RateiDecep
-3.9222.544975.46Apr 27, 2026

Evaluation results

MethodLinks
iDecepAttack Model=Qwen-Plus, Victim Model=Gemini-2.5-Flash2026.0498980
iDecepAttack Model=Qwen-Plus, Victim Model=GPT-4o2026.0496960
iDecepAttack Model=GPT-3.5-Turbo, Victim Model=Gemini-2.5-Flash2026.0491910
iDecepAttack Model=GPT-3.5-Turbo, Victim Model=GPT-4o2026.0487870
iDecepAttack Model=GPT-3.5-Turbo, Victim Model=GPT-52026.04791960
iDecepAttack Model=Qwen-Plus, Victim Model=GPT-52026.04631251
iDecepAttack Model=Qwen-Plus, Victim Model=Claude-Sonnet-4.52026.04592534
Chain of AttackAttack Model=Qwen-Plus, Victim Model=Gemini-2.5-Flash2026.0444440
CrescendoAttack Model=Qwen-Plus, Victim Model=GPT-4o2026.0443430
CrescendoAttack Model=GPT-3.5-Turbo, Victim Model=Gemini-2.5-Flash2026.0443430
iDecepAttack Model=GPT-3.5-Turbo, Victim Model=Claude-Sonnet-4.52026.04362313
CrescendoAttack Model=Qwen-Plus, Victim Model=Gemini-2.5-Flash2026.0435350
Chain of AttackAttack Model=Qwen-Plus, Victim Model=GPT-4o2026.0434340
CrescendoAttack Model=GPT-3.5-Turbo, Victim Model=GPT-4o2026.0431310
Chain of AttackAttack Model=Qwen-Plus, Victim Model=GPT-52026.04202
CrescendoAttack Model=Qwen-Plus, Victim Model=GPT-52026.04202
Chain of AttackAttack Model=Qwen-Plus, Victim Model=Claude-Sonnet-4.52026.04000
Chain of AttackAttack Model=GPT-3.5-Turbo, Victim Model=GPT-4o2026.04000
Chain of AttackAttack Model=GPT-3.5-Turbo, Victim Model=Gemini-2.5-Flash2026.04000
Chain of AttackAttack Model=GPT-3.5-Turbo, Victim Model=Claude-Sonnet-4.52026.04000
Chain of AttackAttack Model=GPT-3.5-Turbo, Victim Model=GPT-52026.04000
CrescendoAttack Model=Qwen-Plus, Victim Model=Claude-Sonnet-4.52026.04000
CrescendoAttack Model=GPT-3.5-Turbo, Victim Model=Claude-Sonnet-4.52026.04000
CrescendoAttack Model=GPT-3.5-Turbo, Victim Model=GPT-52026.04000