List your company

Safety Monitoring on Jailbroken

32.96ASR @ 1% FPRActProbe
30.278448.379266.4884.5808Mar 24, 2026

Evaluation results

MethodLinks
ActProbe2026.0332.960.9653
ActWM2026.0345.920.9292
QwenGuard2026.0361.220.9261
LlamaGuard2026.0364.290.7233
Base2026.03100—