SCIENCE@home

Stochastic Self-Guidance for Training-Free Enhancement of Diffusion Models

Chubin Chen,Jiashu Zhu,Xiaokun Feng,Nisha Huang,Chen Zhu,Meiqi Wu,Fangyuan Mao,Jiahong Wu,Xiangxiang Chu,Xiu Li

Published 2025 in Unknown venue

ABSTRACT

Classifier-free Guidance (CFG) is a widely used technique in modern diffusion models for enhancing sample quality and prompt adherence. However, through an empirical analysis on Gaussian mixture modeling with a closed-form solution, we observe a discrepancy between the suboptimal results produced by CFG and the ground truth. The model's excessive reliance on these suboptimal predictions often leads to semantic incoherence and low-quality outputs. To address this issue, we first empirically demonstrate that the model's suboptimal predictions can be effectively refined using sub-networks of the model itself. Building on this insight, we propose S$^2$-Guidance, a novel method that leverages stochastic block-dropping during the forward process to construct stochastic sub-networks, effectively guiding the model away from potential low-quality predictions and toward high-quality outputs. Extensive qualitative and quantitative experiments on text-to-image and text-to-video generation tasks demonstrate that S$^2$-Guidance delivers superior performance, consistently surpassing CFG and other advanced guidance strategies. Our code will be released.

PUBLICATION RECORD

  • Publication year

    2025

  • Venue

    Unknown venue

  • Publication date

    2025-08-18

  • Fields of study

    Computer Science

  • Identifiers
    arXiv 2508.12880
  • External record

    Open on Semantic Scholar

  • Source metadata

    Semantic Scholar

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-84 of 84 references · Page 1 of 1

CITED BY

Showing 1-19 of 19 citing papers · Page 1 of 1