Localized Schr\"odinger Bridge Sampler

TL;DR

This paper introduces a localized Schr"odinger bridge sampling method that reduces high-dimensional sampling problems into multiple low-dimensional problems, improving efficiency and extending to conditional and Bayesian sampling.

Contribution

It proposes a novel localization strategy for Schr"odinger bridge samplers that exploits conditional independence, enabling scalable high-dimensional sampling and connections to transformer architectures.

Findings

Effective in high-dimensional Gaussian sampling

Demonstrates success on stochastic process data

Extends to conditional and Bayesian inference

Abstract

We consider the problem of sampling from an unknown distribution for which only a sufficiently large number of training samples are available. In this paper, we build on previous work combining Schr\"odinger bridges and plug & play Langevin samplers. A key bottleneck of these approaches is the exponential dependence of the required training samples on the dimension, $d$, of the ambient state space. We propose a localization strategy which exploits conditional independence of conditional expectation values. Localization thus replaces a single high-dimensional Schr\"odinger bridge problem by $d$ low-dimensional Schr\"odinger bridge problems over the available training samples. In this context, a connection to multi-head self attention transformer architectures is established. As for the original Schr\"odinger bridge sampling approach, the localized sampler is stable and geometric ergodic. The sampler also naturally extends to conditional sampling and to Bayesian inference. We demonstrate the performance of our proposed scheme through experiments on a high-dimensional Gaussian problem, on a temporal stochastic process, and on a stochastic subgrid-scale parametrization conditional sampling problem. We also extend the idea of localization to plug & play Langevin samplers using kernel-based denoising in combination with Tweedie's formula.