Where Concept Erasure Should Occur: Concept-Layer Alignment in Text-to-Video Diffusion Models 事件

Where Concept Erasure Should Occur: Concept-Layer Alignment in Text-to-Video Diffusion Models arXiv:2605.25941v1 Announce Type: new Abstract: Text-to-video diffusion transformers encode semantic information unevenly across model depth, which constrains effective concept erasure. We identify a representational bottleneck, termed concept-layer topological alignment, under which target concepts exhibit higher separability at certain representational depths. Outside these depths, concept and non-ta