AI RESEARCH

Beyond Hate: Differentiating Uncivil and Intolerant Speech in Multimodal Content Moderation

arXiv CS.CL

ArXi:2603.22985v1 Announce Type: new Current multimodal toxicity benchmarks typically use a single binary hatefulness label. This coarse approach conflates two fundamentally different characteristics of expression: tone and content. Drawing on communication science theory, we