NETS: A Non-Equilibrium Transport Sampler

TL;DR

NETS is a novel non-equilibrium sampling algorithm that improves upon annealed importance sampling by incorporating a learned drift term, leading to unbiased estimates and better performance on complex distributions.

Contribution

The paper introduces NETS, a new sampling method that leverages learned drifts and unbiased objective functions to enhance sampling efficiency and accuracy.

Findings

NETS is unbiased and tunable for effective sample size.

It outperforms existing methods on benchmarks and high-dimensional distributions.

Demonstrates success on statistical lattice field theory models.

Abstract

We propose an algorithm, termed the Non-Equilibrium Transport Sampler (NETS), to sample from unnormalized probability distributions. NETS can be viewed as a variant of annealed importance sampling (AIS) based on Jarzynski's equality, in which the stochastic differential equation used to perform the non-equilibrium sampling is augmented with an additional learned drift term that lowers the impact of the unbiasing weights used in AIS. We show that this drift is the minimizer of a variety of objective functions, which can all be estimated in an unbiased fashion without backpropagating through solutions of the stochastic differential equations governing the sampling. We also prove that some these objectives control the Kullback-Leibler divergence of the estimated distribution from its target. NETS is shown to be unbiased and, in addition, has a tunable diffusion coefficient which can be adjusted post-training to maximize the effective sample size. We demonstrate the efficacy of the method on standard benchmarks, high-dimensional Gaussian mixture distributions, and a model from statistical lattice field theory, for which it surpasses the performances of related work and existing baselines.