Paper
Publication
Continual Repeated Annealed Flow Transport Monte Carlo
Abstract

In this paper, we propose Continual Repeated Annealed Flow Transport Monte Carlo (CRAFT), a method that combines a sequential Monte Carlo (SMC) sampler (itself a generalization of Annealed Importance Sampling) with variational inference using normalizing flows. The normalizing flows are directly trained to transport between annealing temperatures using a KL divergence for each transition. This optimization objective is itself estimated using the normalizing flow/SMC approximation. We show that CRAFT conceptually and empirically improves on Annealed Flow Transport Monte Carlo (Arbel et al., 2021), on which it builds and also on MCMC based Stochastic Normalizing Flows (Wu et al., 2020). By combining CRAFT with particle Markov chain Monte Carlo, we show that such learnt samplers can achieve impressively accurate results on a challenging lattice field theory example.

Authors' notes