Entropic Independence II: Optimal Sampling and Concentration via Restricted Modified Log-Sobolev Inequalities

TL;DR

This paper introduces a new framework using restricted modified log-Sobolev inequalities to achieve nearly-linear mixing times for sampling from complex models like the hardcore and Ising models, even on high-degree graphs.

Contribution

It develops a novel approach to establish entropy contraction near stationarity, leading to optimal mixing time bounds independent of maximum degree.

Findings

Achieves nearly-linear time sampling for hardcore and Ising models with a constant gap to the tree-uniqueness threshold.

Improves prior mixing time bounds that depended on maximum degree or were quadratic in n.

Extends entropic independence to spectrally independent distributions under restricted external fields.

Abstract

We introduce a framework for obtaining tight mixing times for Markov chains based on what we call restricted modified log-Sobolev inequalities. Modified log-Sobolev inequalities (MLSI) quantify the rate of relative entropy contraction for the Markov operator, and are notoriously difficult to establish. However, infinitesimally close to stationarity, entropy contraction becomes equivalent to variance contraction, a.k.a. a Poincare inequality, which is significantly easier to establish through, e.g., spectral analysis. Motivated by this observation, we study restricted modified log-Sobolev inequalities that guarantee entropy contraction not for all starting distributions, but for those in a large neighborhood of the stationary distribution. We show how to sample from the hardcore and Ising models on $n$-node graphs that have a constant $\delta$ relative gap to the tree-uniqueness threshold, in nearly-linear time $\widetilde O_{\delta}(n)$. Notably, our bound does not depend on the maximum degree $\Delta$, and is therefore optimal even for high-degree graphs. This improves on prior mixing time bounds of $\widetilde O_{\delta, \Delta}(n)$ and $\widetilde O_{\delta}(n^2)$, established via (non-restricted) modified log-Sobolev and Poincare inequalities respectively. We further show that optimal concentration inequalities can still be achieved from the restricted form of modified log-Sobolev inequalities. To establish restricted entropy contraction, we extend the entropic independence framework of Anari, Jain, Koehler, Pham, and Vuong to the setting of distributions that are spectrally independent under a restricted set of external fields. We also develop an orthogonal trick that might be of independent interest: utilizing Bernoulli factories we show how to implement Glauber dynamics updates on high-degree graphs in $O(1)$ time, assuming standard adjacency array representation of the graph.