List your company

Toxic Comment Detection on Toxic Comment

47.3ASRCriteria Attack
-0.43611.95724.3536.743Jan 15, 2026

Evaluation results

MethodLinks
Criteria AttackDefense=StruQ2026.0147.3
CombinedDefense=StruQ2026.0118.1
TopicDefense=SecAlign2026.0115.4
Criteria AttackDefense=SecAlign2026.0115.2
IgnoreDefense=StruQ2026.0114.2
TopicDefense=StruQ2026.0112.3
Escape SeparationDefense=StruQ2026.0112.1
Fake CompletionDefense=StruQ2026.0111.9
IgnoreDefense=SecAlign2026.019.1
Separator InjectionDefense=StruQ2026.018.8
Escape SeparationDefense=SecAlign2026.014.7
CombinedDefense=SecAlign2026.012.1
Fake CompletionDefense=SecAlign2026.011.4
Separator InjectionDefense=SecAlign2026.011.4