List your company

Jailbreak Attack on ShadowRisk

100ASR-KWPAIR
67.7676.1384.592.87Jan 7, 2026

Evaluation results

MethodLinks
PAIRVictim Model=Gemini-3-Flash, Search Setting=w/ agentic2026.0110052
FlipAttackVictim Model=DeepSeek-V3.2, Search Setting=w/ snippet2026.0110023
SearchAttackVictim Model=DeepSeek-V3.2, Search Setting=w/ snippet2026.0110015
SearchAttackVictim Model=Qwen3-32B, Search Setting=w/ snippet2026.0110027
SearchAttackVictim Model=GPT-4o, Search Setting=w/ snippet2026.0110015
SearchAttackVictim Model=Gemini-3-Flash, Search Setting=w/ snippet2026.019971
SearchAttackVictim Model=Gemini-3-Flash, Search Setting=w/ agentic2026.019975
FlipAttackVictim Model=DeepSeek-V3.2, Search Setting=w/ agentic2026.019928
PAIRVictim Model=Qwen3-32B, Search Setting=w/ agentic2026.019923
SearchAttackVictim Model=Qwen3-32B, Search Setting=w/ agentic2026.019933
SearchAttackVictim Model=GPT-4o, Search Setting=w/ agentic2026.019920
PAIRVictim Model=Gemini-3-Flash, Search Setting=w/ snippet2026.019848
ReNeLLMVictim Model=Gemini-3-Flash, Search Setting=w/ snippet2026.01980
ReNeLLMVictim Model=Gemini-3-Flash, Search Setting=w/ agentic2026.01980
FlipAttackVictim Model=Gemini-3-Flash, Search Setting=w/ snippet2026.019863
ReNeLLMVictim Model=DeepSeek-V3.2, Search Setting=w/ snippet2026.01980
SearchAttackVictim Model=DeepSeek-V3.2, Search Setting=w/ agentic2026.019834
PAIRVictim Model=Qwen3-32B, Search Setting=w/ snippet2026.019817
ReNeLLMVictim Model=Qwen3-32B, Search Setting=w/ snippet2026.01980
FlipAttackVictim Model=Gemini-3-Flash, Search Setting=w/ agentic2026.019768
SearchAttackVictim Model=Gemini-3-Flash, Search Setting=w/o search2026.019763
SearchAttackVictim Model=DeepSeek-V3.2, Search Setting=w/o search2026.01970
FlipAttackVictim Model=Qwen3-32B, Search Setting=w/ agentic2026.019732
PAIRVictim Model=GPT-4o, Search Setting=w/ snippet2026.01978
PAIRVictim Model=GPT-4o, Search Setting=w/ agentic2026.019713
ReNeLLMVictim Model=GPT-4o, Search Setting=w/ snippet2026.01970
PAIRVictim Model=Gemini-3-Flash, Search Setting=w/o search2026.019637
FlipAttackVictim Model=Gemini-3-Flash, Search Setting=w/o search2026.019652
ReNeLLMVictim Model=Qwen3-32B, Search Setting=w/ agentic2026.01960
FlipAttackVictim Model=Qwen3-32B, Search Setting=w/ snippet2026.019621
ReNeLLMVictim Model=Gemini-3-Flash, Search Setting=w/o search2026.01950
ReNeLLMVictim Model=DeepSeek-V3.2, Search Setting=w/ agentic2026.01950
FlipAttackVictim Model=DeepSeek-V3.2, Search Setting=w/o search2026.01951
ReNeLLMVictim Model=GPT-4o, Search Setting=w/ agentic2026.01950
SearchAttackVictim Model=Qwen3-32B, Search Setting=w/o search2026.019414
PAIRVictim Model=DeepSeek-V3.2, Search Setting=w/ snippet2026.01939
FlipAttackVictim Model=GPT-4o, Search Setting=w/ snippet2026.019312
FlipAttackVictim Model=GPT-4o, Search Setting=w/ agentic2026.019317
SearchAttackVictim Model=GPT-4o, Search Setting=w/o search2026.01938
ReNeLLMVictim Model=GPT-4o, Search Setting=w/o search2026.01910
PAIRVictim Model=GPT-4o, Search Setting=w/o search2026.01882
PAIRVictim Model=Qwen3-32B, Search Setting=w/o search2026.01878
FlipAttackVictim Model=Qwen3-32B, Search Setting=w/o search2026.018712
PAIRVictim Model=DeepSeek-V3.2, Search Setting=w/ agentic2026.018414
ReNeLLMVictim Model=Qwen3-32B, Search Setting=w/o search2026.01830
PAIRVictim Model=DeepSeek-V3.2, Search Setting=w/o search2026.01760
ReNeLLMVictim Model=DeepSeek-V3.2, Search Setting=w/o search2026.01740
FlipAttackVictim Model=GPT-4o, Search Setting=w/o search2026.01695