Abstract:There has been a recent surge of powerful tools to show rapid mixing of Markov chains, via functional inequalities such as Poincaré inequalities. In many situations, Markov chains fail to mix rapidly from a worst-case initialization, yet are expected to approximately sample from a random initialization. For example, this occurs if the target distribution has metastable states, small clusters accounting for a vanishing fraction of the mass that are essentially disconnected from the bulk of the measure. Under such conditions, a Poincaré inequality cannot hold, necessitating new tools to prove sampling guarantees. We develop a framework to analyze simulated annealing, based on establishing so-called weak Poincaré inequalities. These inequalities imply mixing from a suitably warm start, and simulated annealing provides a way to chain such warm starts together into a sampling algorithm. We further identify a local-to-global principle to prove weak Poincaré inequalities, mirroring the spectral independence and localization schemes frameworks for analyzing mixing times of Markov chains. As our main application, we prove that simulated annealing samples from the Gibbs measure of a spherical spin glass for inverse temperatures up to a natural threshold, matching recent algorithms based on algorithmic stochastic localization. This provides the first Markov chain sampling guarantee that holds beyond the uniqueness threshold for spherical spin glasses, where mixing from a worst-case initialization is provably slow. As an ingredient in our proof, we prove bounds on the operator norm of the covariance matrix of spherical spin glasses in the full replica-symmetric regime. Additionally, we resolve questions related to the mixing of Glauber dynamics in the ferromagnetic Potts model from a uniform monochromatic coloring, and sampling using data-based initializations.

Sampling from the Sherrington-Kirkpatrick Gibbs measure via algorithmic stochastic localization

Sampling from Mean-Field Gibbs Measures via Diffusion Processes

Sampling from Spherical Spin Glasses in Total Variation via Algorithmic Stochastic Localization

Weak Poincaré Inequalities, Simulated Annealing, and Sampling from Spherical Spin Glasses

Global Optimization Via Schrödinger-Föllmer Diffusion.

Hardness of sampling for the anti-ferromagnetic Ising model on random graphs

Sampling Approximately Low-Rank Ising Models: MCMC meets Variational Methods

Perfect Sampling in Infinite Spin Systems Via Strong Spatial Mixing

Disordered Gibbs measures and Gaussian conditioning

Stochastic Quantum Sampling for Non-Logconcave Distributions and Estimating Partition Functions

On sampling two spin models using the local connective constant

Sampling from Potts on random graphs of unbounded degree via random-cluster dynamics

A Near-Linear Time Sampler for the Ising Model with External Field

Learning the Sherrington-Kirkpatrick Model Even at Low Temperature

What Can Be Sampled Locally?

Concentration of the Langevin Algorithm's Stationary Distribution

Spin glasses and Stein's method

Joint parameter estimations for spin glasses

Global Optimization Algorithm through High-Resolution Sampling

Tamed Langevin sampling under weaker conditions

TAP free energy, spin glasses, and variational inference