List your company

Safety Monitoring on Multilingual

34.15ASR @ 1% FPRActWM
31.51649.295567.07584.8545Mar 24, 2026

Evaluation results

MethodLinks
ActWM2026.0334.1596.19
QwenGuard2026.0339.0292.27
ActProbe2026.0362.1494.84
LlamaGuard2026.0365.8543.53
Base2026.03100—