Representative Guidance: Diffusion Model Sampling with Coherence

Anh-Dung Dinh, Daochang Liu, Chang Xu

Keywords: generative models, diffusion model

TL;DR: The paper investigates the inconsistency property of the diffusion model and propose a guidance method based on representative information to fix the inconsistency problem

Abstract:

The diffusion sampling process faces a persistent challenge stemming from its incoherence, attributable to varying noise directions across different timesteps.
Our Representative Guidance (RepG) offers a new perspective to address this issue by reformulating the sampling process with a coherent direction toward a representative target.
From this perspective, classic classifier guidance reveals its drawback in lacking meaningful representative information, as the features it relies on are optimized for discrimination and tend to highlight only a narrow set of class-specific cues. This focus often sacrifices diversity and increases the risk of adversarial generation.
In contrast, we leverage self-supervised representations as the coherent target and treat sampling as a downstream task—one that focuses on refining image details and correcting generation errors, rather than settling for oversimplified outputs.
Our Representative Guidance achieves superior performance and demonstrates the potential of pre-trained self-supervised models in guiding diffusion sampling. Our findings show that RepG not only significantly improves vanilla diffusion sampling, but also surpasses state-of-the-art benchmarks when combined with classifier-free guidance.

Primary Area: generative models

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 8945