List your company

Safety Monitoring on AutoDAN

67.86ASR @ 1% FPRActWM
66.574475.252283.9392.6078Mar 24, 2026

Evaluation results

MethodLinks
ActWM2026.0367.860.9048
ActProbe2026.0375.320.8746
QwenGuard2026.0387.50.7851
LlamaGuard2026.0389.880.4092
Base2026.03100—