List your company

Jailbreak Attack on StrategyQA (ASR, HS)

3Attack Success Rate (ASR)H-COT
-0.3222.0944.566.91Apr 17, 2026

Evaluation results

MethodLinks
H-COTVictim LRM=Qwen2.5-Max2026.0432.33
H-COTVictim LRM=OpenAI o4-mini2026.0442
Cognitive OverloadVictim LRM=Qwen2.5-Max2026.0462
Cognitive OverloadVictim LRM=OpenAI o4-mini2026.04202.05
H-COTVictim LRM=DeepSeek R12026.04222.82
Cognitive OverloadVictim LRM=DeepSeek R12026.04282.5
PRJAVictim LRM=OpenAI o4-mini2026.04782.76
PRJAVictim LRM=Qwen2.5-Max2026.04842.93
PRJAVictim LRM=DeepSeek R12026.04862.97