English

Learning Diffusion Models with Flexible Representation Guidance

Machine Learning 2025-10-14 v2 Artificial Intelligence Computer Vision and Pattern Recognition

Abstract

Diffusion models can be improved with additional guidance towards more effective representations of input. Indeed, prior empirical work has already shown that aligning internal representations of the diffusion model with those of pre-trained models improves generation quality. In this paper, we present a systematic framework for incorporating representation guidance into diffusion models. We provide alternative decompositions of denoising models along with their associated training criteria, where the decompositions determine when and how the auxiliary representations are incorporated. Guided by our theoretical insights, we introduce two new strategies for enhancing representation alignment in diffusion models. First, we pair examples with target representations either derived from themselves or arisen from different synthetic modalities, and subsequently learn a joint model over the multimodal pairs. Second, we design an optimal training curriculum that balances representation learning and data generation. Our experiments across image, protein sequence, and molecule generation tasks demonstrate superior performance as well as accelerated training. In particular, on the class-conditional ImageNet 256×256256\times 256 benchmark, our guidance results in 23.323.3 times faster training than the original SiT-XL as well as four times speedup over the state-of-the-art method REPA. The code is available at https://github.com/ChenyuWang-Monica/REED.

Keywords

Cite

@article{arxiv.2507.08980,
  title  = {Learning Diffusion Models with Flexible Representation Guidance},
  author = {Chenyu Wang and Cai Zhou and Sharut Gupta and Zongyu Lin and Stefanie Jegelka and Stephen Bates and Tommi Jaakkola},
  journal= {arXiv preprint arXiv:2507.08980},
  year   = {2025}
}

Comments

NeurIPS 2025; Also Oral at ICML 2025 FM4LS workshop

R2 v1 2026-07-01T03:57:21.676Z