Where Should a Diffusion Transformer Meet Its Teacher?
Representation Alignment (REPA) significantly accelerates Diffusion Transformer training by guiding intermediate student features with a pretrained visual teacher, yet existing methods typically attach supervision to a single fixed layer. As student representations evolve, however, we find that the most effective layer for alignment varies across training stages. To track this evolving layer suitability online without costly search, we measure gradient compatibility between the flow-matching and alignment objectives, and propose AdaREPA, an online routing framework that adaptively selects the supervision layer as training progresses. AdaREPA uses a softmax distribution over smoothed layer compatibility scores to sample a layer for alignment at each training step. This probabilistic routing adapts teacher supervision to the student's changing learning needs, with minimal training overhead and no additional inference cost. On ImageNet-256, AdaREPA accelerates convergence and improves generation quality across model sizes, while revealing how the most suitable layers for supervision shift over training.
Authors
- Yihao Ouyang (ORCID: https://orcid.org/0009-0007-8315-2139)
- Zhuoqi Hu (ORCID: https://orcid.org/0009-0006-9446-8465)
- Ang Wang
- Jiamang Wang
- Wenting Shen
- Minyi Guo
- Shiwei Li
- Haozhao Wang
- Xiandi Luo
- Ruixuan Li
Publication Details
- Journal
- arXiv (Cornell University)
- Published
- 2026-11-18
- DOI
- https://doi.org/10.48550/arxiv.2611.24380
- Primary Topic
- Generative Adversarial Networks and Image Synthesis
- Type
- preprint