arXiv cs.AI· Wei Guo, Yaowen Zhang, Xingtong Ge, Jun Zhang·· 9 天前AI 评分22
扩散模型中的扰动传播：相同路径代价为何产生不同输出影响
Equal Path Cost, Unequal Output Effects: Understanding Perturbation Propagation in Diffusion Models
AI 导读
研究提出一个理论框架分析扩散模型采样中扰动的传播机制，用扰动与参考轨迹分布间的 KL 散度定义"路径代价"，证明其能约束但不决定输出分布的变化。在预训练扩散模型上以相同路径代价进行受控干预，发现输出敏感度随采样阶段和空间频率呈现不同模式；将该分析用于缓存加速时，能可靠识别缓存导致较大图像误差的采样区间。
正文 · 原文
Abstract:Diffusion models have achieved remarkable success in generative modeling, with their sampling procedures routinely modified to control generation and improve efficiency. These modifications introduce perturbations along the sampling trajectory, raising a central question: how do such perturbations affect generated output? To address this question, we develop a theoretical framework to investigate perturbation propagation, combining dynamical analysis of the sampling process with an information-theoretic characterization of output responses. Within this framework, we quantify perturbation strength using the Kullback--Leibler (KL) divergence between perturbed and reference trajectory distributions, termed as path cost, which is shown to bound, but do not determine, changes in the output distribution. Building on this analysis, we derive a response identity that connects the propagation and accumulation of local perturbations with the information captured by a selected feature mean, explaining why changes in the output distribution can remain undetected by its first-order response. We test our theoretical analysis through controlled interventions at equal path cost in pretrained diffusion models, revealing distinct patterns of output sensitivity across sampling stages and spatial frequencies. To assess whether our framework can diagnose perturbations arising from practical approximations, we apply it to cache-based acceleration and show that our propagation analysis reliably identifies sampling intervals where caching causes larger image errors.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.11380 [cs.AI] |
| (or arXiv:2610.11380v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.11380 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Wei Guo [view email]
[v1]
Thu, 8 Oct 2026 07:12:36 UTC (1,304 KB)
来源：arXiv cs.AI · arxiv.org