MLLGCOFeb 16, 2024

Stochastic Localization via Iterative Posterior Sampling

arXiv:2402.10758v229 citationsh-index: 9ICML
AI Analysis

This work provides a novel sampling technique for complex distributions, addressing a key challenge in Bayesian inference and statistical mechanics, though it builds incrementally on existing score-based learning frameworks.

The authors tackled the problem of sampling from unnormalized target densities by introducing SLIPS, a stochastic localization method with iterative posterior sampling, achieving effective sampling on multi-modal benchmarks like Gaussian mixtures and high-dimensional field systems.

Building upon score-based learning, new interest in stochastic localization techniques has recently emerged. In these models, one seeks to noise a sample from the data distribution through a stochastic process, called observation process, and progressively learns a denoiser associated to this dynamics. Apart from specific applications, the use of stochastic localization for the problem of sampling from an unnormalized target density has not been explored extensively. This work contributes to fill this gap. We consider a general stochastic localization framework and introduce an explicit class of observation processes, associated with flexible denoising schedules. We provide a complete methodology, $\textit{Stochastic Localization via Iterative Posterior Sampling}$ (SLIPS), to obtain approximate samples of this dynamics, and as a by-product, samples from the target distribution. Our scheme is based on a Markov chain Monte Carlo estimation of the denoiser and comes with detailed practical guidelines. We illustrate the benefits and applicability of SLIPS on several benchmarks of multi-modal distributions, including Gaussian mixtures in increasing dimensions, Bayesian logistic regression and a high-dimensional field system from statistical-mechanics.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes