List your company

Attack Success Rate Evaluation on SKILL-INJECT Contextual

0Attack Success Rate (ASR)Direct Attack
-3.5220.244467.76Apr 5, 2026

Evaluation results

MethodLinks
Direct AttackModel=kimi-k2.52026.040
Direct AttackModel=claude-sonnet-4-5-202509292026.042
Direct AttackModel=Qwen3.5-Plus2026.042
Direct AttackModel=glm-52026.042
Direct AttackModel=MiniMax-M2.52026.044
Direct AttackModel=gpt-5.42026.045
SKILL-INJECTModel=kimi-k2.52026.045
SKILL-INJECTModel=Qwen3.5-Plus2026.045
Direct AttackModel=gemini-3.0-pro-preview2026.047
SKILL-INJECTModel=MiniMax-M2.52026.047
Direct AttackModel=gpt-5.4-nano2026.0410
SKILL-INJECTModel=gemini-3.0-pro-preview2026.0410
Direct AttackModel=doubao-seed-1-8-2512282026.0410
SKILL-INJECTModel=gpt-5.4-nano2026.0412
Direct AttackModel=hunyuan-2.0-instruct-202511112026.0412
SKILL-INJECTModel=glm-52026.0415
SKILL-INJECTModel=doubao-seed-1-8-2512282026.0415
SKILL-INJECTModel=gpt-5.42026.0419
SKILL-INJECTModel=claude-sonnet-4-5-202509292026.0420
SKILL-INJECTModel=hunyuan-2.0-instruct-202511112026.0420
SkillAttackModel=Qwen3.5-Plus2026.0456
SkillAttackModel=claude-sonnet-4-5-202509292026.0464
SkillAttackModel=gemini-3.0-pro-preview2026.0468
SkillAttackModel=glm-52026.0468
SkillAttackModel=gpt-5.42026.0471
SkillAttackModel=doubao-seed-1-8-2512282026.0471
SkillAttackModel=gpt-5.4-nano2026.0476
SkillAttackModel=MiniMax-M2.52026.0478
SkillAttackModel=kimi-k2.52026.0485
SkillAttackModel=hunyuan-2.0-instruct-202511112026.0488