Representation by Design in Generation: Cross-View Class-Token Alignment in Diffusion Transformers
arXiv:2609.36348v1 Announce Type: new Abstract: Generative and representation learning remain asymmetrically connected: semantic representations are used to improve diffusion generation, whereas the models' own representations are often treated as a by-product of synthesis. We ask whether diffusion…
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-09-30 04:00 · arXiv cs.CV
Representation by Design in Generation: Cross-View Class-Token Alignment in Diffusion Transformers