RBRC seminar

Abstract: A class of deep generative models called Normalizing Flows has been recently proposed as a promising alternative to conventional Markov Chain Monte Carlo simulations to sample lattice field theory configurations: these models provide a unique approach to potentially avoid the large autocorrelations that characterize Monte Carlo simulations close to the continuum limit. In this talk we explore the novel concept of Stochastic Normalizing Flows (SNFs), in which neural-network layers are combined with traditional Monte Carlo updates: in particular, we show how SNFs share the same theoretical framework of out-of-equilibrium simulations based on Jarzynski's equality. The latter is a well-known result in non-equilibrium statistical mechanics that has already been successfully used in computations of free-energy differences in lattice gauge theories. We discuss how the connection between Normalizing Flows and Jarzynski's equality can be exploited to optimize the efficiency of this extended class of generative models and we present some numerical results in an example of application.