Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.25941 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917531771469824 |
|---|---|
| author | Xie, Yiwei Liu, Ping Zhang, Zheng |
| author_facet | Xie, Yiwei Liu, Ping Zhang, Zheng |
| contents | 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-target signals remain strongly entangled, limiting the effectiveness of depth-specific erasure. This observation reframes concept erasure as the problem of identifying representational depths where concept-non-target separation naturally emerges. Motivated by this structural constraint, we introduce CLEAR, a separability-driven optimization framework for concept erasure that explicitly enforces concept-layer alignment. CLEAR operationalizes this principle by formulating layer selection as an optimization problem over concept-non-target separability, rather than relying on layer-agnostic or heuristic choices. To enable this, we introduce a separability-aware objective that favors layers exhibiting stronger concept-non-target separation. Experiments on large-scale text-to-video diffusion models demonstrate that enforcing concept--layer alignment leads to more precise concept suppression while preserving overall generative quality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_25941 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Where Concept Erasure Should Occur: Concept-Layer Alignment in Text-to-Video Diffusion Models Xie, Yiwei Liu, Ping Zhang, Zheng Computer Vision and Pattern Recognition 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-target signals remain strongly entangled, limiting the effectiveness of depth-specific erasure. This observation reframes concept erasure as the problem of identifying representational depths where concept-non-target separation naturally emerges. Motivated by this structural constraint, we introduce CLEAR, a separability-driven optimization framework for concept erasure that explicitly enforces concept-layer alignment. CLEAR operationalizes this principle by formulating layer selection as an optimization problem over concept-non-target separability, rather than relying on layer-agnostic or heuristic choices. To enable this, we introduce a separability-aware objective that favors layers exhibiting stronger concept-non-target separation. Experiments on large-scale text-to-video diffusion models demonstrate that enforcing concept--layer alignment leads to more precise concept suppression while preserving overall generative quality. |
| title | Where Concept Erasure Should Occur: Concept-Layer Alignment in Text-to-Video Diffusion Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.25941 |