List your company

Toxicity Mitigation on DailyDialog

1.16RTR (Backdoor Non-toxic)Attack w/o defense
-0.02560.28220.590.8978Jul 8, 2025

Evaluation results

MethodLinks
Attack w/o defenseDefense setting=Attack w/o defense2025.071.1662.1813.7654.78
Multi-level filterDefense setting=Multi-level filter2025.070.1615.561.718.08
Optimus (CH)Defense setting=Optimus (CH)2025.070.020.040.020.64