SoftHateBench: Evaluating Moderation Models Against Reasoning-Driven, Policy-Compliant Hostility.

Xuanyu Su, Diana Inkpen, Nathalie Japkowicz

Browse the full WWW paper archive.