arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2607.04442stat.MLcs.LG
收藏

收紧扩散模型的得分匹配间隙

Tightening the Score Matching Gap for Diffusion Models

Benjamin Dupuis, Tyler Farghly, Maxime Haddouche, Alain Durmus, Umut Simsekli

AI总结:

针对扩散模型的得分匹配间隙问题,该研究利用得分估计器正则性与反向过程收缩性,推导三类指标的更紧边界,揭示低噪声尺度下得分近似质量对缩窄间隙的关键作用。

摘要:

扩散模型(DMs)是一类用于从未知分布中近似采样的前沿生成方法。其训练与评估主要依赖证据下界(ELBO),该界将模型样本的KL散度与路径上的得分匹配损失关联起来,后者作为可处理的替代指标。由该界导出的样本质量与得分匹配损失之间的差异被称为得分匹配间隙,已知该间隙在最坏情况下是紧的,但通常无法准确描述样本质量。本工作对该间隙开展理论分析,有效利用得分估计器类别的正则性,针对KL散度、反向KL散度与Wasserstein距离三类指标推导了更紧的边界。研究结果表明,在低噪声尺度下,得分近似的质量对缩窄得分匹配间隙的影响更大。为推导上述边界,核心技术思路是利用反向过程的收缩性质,具体而言,借助熵流、对数Sobolev不等式与反射耦合,将Langevin扩散的遍历性与得分匹配间隙问题严格关联起来。

英文摘要:

Diffusion models (DMs) are a state-of-the-art generative method to approximately sample from an unknown distribution. Their training and evaluation primarily rely on an Evidence Lower Bound (ELBO), which relates the Kullback-Leibler (KL) divergence of model samples to the score matching loss along the path, which serves as a tractable surrogate. The difference between sample quality and the score matching loss produced by this bound leads to the \emph{score matching gap}, which is known to be tight in the worst-case but not descriptive of sample quality in general. In this work, we provide a theoretical analysis of this gap, developing tighter bounds for three metrics: KL divergence, reverse KL divergence, and Wasserstein distance, effectively exploiting the regularity of the class of score estimators. Our results suggest that the quality of the score approximation has more impact on closing the score matching gap for low noise scales. To obtain these bounds, our key technical insight is to exploit the contraction properties of the backward processes. In particular, we rely on entropy flows, logarithmic Sobolev inequalities and reflection couplings, rigorously linking the ergodicity of the Langevin diffusion to the score matching gap problem.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑