Fast Convergence of Langevin Dynamics on Manifold: Geodesics meet Log-Sobolev

Abstract: Sampling is a fundamental and arguably very important task with numerous applications in Machine Learning. One approach to sample from a high dimensional distribution $e<sup>{-f}$ for some function $f$ is the Langevin Algorithm (LA). Recently, there has been a lot of progress in showing fast convergence of LA even in cases where $f$ is non-convex, notably [53], [39] in which the former paper focuses on functions $f$ defined in $\mathbb{R}<sup>n$ and the latter paper focuses on functions with symmetries (like matrix completion type objectives) with manifold structure. Our work generalizes the results of [53] where $f$ is defined on a manifold $M$ rather than $\mathbb{R}<sup>n$. From technical point of view, we show that KL decreases in a geometric rate whenever the distribution $e<sup>{-f}$ satisfies a log-Sobolev inequality on $M$.