arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2605.08882cs.LG
收藏

离散流匹配:在最小假设下的收敛保证

Discrete Flow Matching: Convergence Guarantees Under Minimal Assumptions

École Polytechnique, CMAP, IP(巴黎理工学院,CMAP,IP);Università degli Studi di Padova(帕多瓦大学);LaMME, Université Évry Paris-Saclay(Évry巴黎萨克雷大学的LaMME)

Le-Tuyet-Nhi Pham, Giovanni Conforti, Zhenjie Ren, Alain Durmus

AI总结:

本文研究了离散流匹配模型在时间离散化下的收敛性,推导了非渐近的KL散度和总变差距离的收敛界,仅依赖于近似误差假设,改进了词汇量和维度的依赖性。

摘要:

流匹配最近作为一种流行的生成模型出现,用于从源分布μ₀采样中模拟目标分布μ₁。该框架依赖于μ₀和μ₁之间的固定耦合,以及确定性或随机桥梁来定义两者之间的插值过程。该过程的时间边缘可以通过估计其马尔可夫投影的转移率或更一般性的生成器来近似采样。该框架最近被扩展到离散源和目标分布的情况,称为离散流匹配(DFM)。然而,此类模型的理论保证仍然稀少。在本文中,我们研究了两个DFM模型在ℤ_m^d = {0,…,m−1}^d上,通过时间离散化进行采样,并推导了这两种模型的非渐近关联界。与之前的工作不同,我们建立了针对目标分布早期停止版本的KL散度的非渐近界。我们还推导了相对于真实目标分布的总变差距离的显式收敛保证。重要的是,这些界仅依赖于近似误差假设,放松了以往工作中使用的标准分数假设,同时也改进了词汇量m和维度d的依赖性。

英文摘要:

Flow Matching has recently emerged as a popular class of generative models for simulating a target distribution $μ_1$ from samples drawn from a source distribution $μ_0$. This framework relies on a fixed coupling between $μ_0$ and $μ_1$, and on a deterministic or stochastic bridge to define an interpolating process between the two distributions. The time marginals of this process can then be approximately sampled by estimating the transition rates, or more generally the generator, of its Markovian projection. This framework has recently been extended to the case of discrete source and target distributions, under the name Discrete Flow Matching (DFM). However, theoretical guarantees for such models remain scarce. In this paper, we study two DFM models on $\mathbb{Z}_m^d = \{0,\ldots,m-1\}^d$, sampled through time discretization, and derive non-asymptotic associated bounds for both of them. In contrast to previous work, we establish non-asymptotic bounds in Kullback--Leibler divergence for the early-stopped version of the target distribution. We also derive explicit convergence guarantees in total variation distance with respect to the true target distribution. Importantly, these bounds rely only on an approximation error assumption, relaxing standard score assumptions used in earlier works, while also yielding improved dependence on the vocabulary size $m$ and the dimension $d$.

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。

↑