Estimation of Thermodynamic Observables in Lattice Field Theories with Deep Generative Models

In this work, we demonstrate that applying deep generative machine learning models for lattice field theory is a promising route solving problems where Markov Chain Monte Carlo (MCMC) methods are problematic. More specifically, show can be used to estimate the absolute value of free energy, which in contrast existing MCMC-based limited only energy differences. We effectiveness proposed method two-dimensional $\phi^4$ and compare it detailed numerical experiments.