Resolving the Mixing Time of the Langevin Algorithm to its Stationary Distribution for Log-Concave Sampling

TL;DR

This paper provides a complete characterization of the mixing time for the Langevin Algorithm in sampling from log-concave distributions, introducing a novel technique from differential privacy to achieve optimal bounds.

Contribution

It introduces a new method using Privacy Amplification by Iteration and Rényi divergence to unify and simplify the analysis of Langevin Algorithm's mixing time.

Findings

Provides tight mixing bounds for Langevin Algorithm in log-concave sampling.

Unifies analysis across various settings including projections and stochastic gradients.

Achieves exponential improvement in mixing time for strongly convex potentials.

Abstract

Sampling from a high-dimensional distribution is a fundamental task in statistics, engineering, and the sciences. A canonical approach is the Langevin Algorithm, i.e., the Markov chain for the discretized Langevin Diffusion. This is the sampling analog of Gradient Descent. Despite being studied for several decades in multiple communities, tight mixing bounds for this algorithm remain unresolved even in the seemingly simple setting of log-concave distributions over a bounded domain. This paper completely characterizes the mixing time of the Langevin Algorithm to its stationary distribution in this setting (and others). This mixing result can be combined with any bound on the discretization bias in order to sample from the stationary distribution of the continuous Langevin Diffusion. In this way, we disentangle the study of the mixing and bias of the Langevin Algorithm. Our key insight is to introduce a technique from the differential privacy literature to the sampling literature. This technique, called Privacy Amplification by Iteration, uses as a potential a variant of R\'enyi divergence that is made geometrically aware via Optimal Transport smoothing. This gives a short, simple proof of optimal mixing bounds and has several additional appealing properties. First, our approach removes all unnecessary assumptions required by other sampling analyses. Second, our approach unifies many settings: it extends unchanged if the Langevin Algorithm uses projections, stochastic mini-batch gradients, or strongly convex potentials (whereby our mixing time improves exponentially). Third, our approach exploits convexity only through the contractivity of a gradient step -- reminiscent of how convexity is used in textbook proofs of Gradient Descent. In this way, we offer a new approach towards further unifying the analyses of optimization and sampling algorithms.