Flow-based sampling for multimodal and extended-mode distributions in lattice field theory
Abstract
Recent results have demonstrated that samplers constructed with flow-based generative models are a promising new approach for configuration generation in lattice field theory. In this paper, we present a set of training- and architecture-based methods to construct flow models for targets with multiple separated modes (i.e.~vacua) as well as targets with extended/continuous modes. We demonstrate the application of these methods to modeling two-dimensional real and complex scalar field theories in their symmetry-broken phases. In this context we investigate different flow-based sampling algorithms, including a composite sampling algorithm where flow-based proposals are occasionally augmented by applying updates using traditional algorithms like HMC.
Cite
@article{arxiv.2107.00734,
title = {Flow-based sampling for multimodal and extended-mode distributions in lattice field theory},
author = {Daniel C. Hackett and Chung-Chun Hsieh and Sahil Pontula and Michael S. Albergo and Denis Boyda and Jiunn-Wei Chen and Kai-Feng Chen and Kyle Cranmer and Gurtej Kanwar and Phiala E. Shanahan},
journal= {arXiv preprint arXiv:2107.00734},
year = {2025}
}
Comments
38+3 pages, 39 figures. v2: major revisions including new application to extended modes