Scaling of Stochastic Normalizing Flows in $\mathrm{SU}(3)$ lattice gauge theory

Abstract

Non-equilibrium Markov Chain Monte Carlo (NE-MCMC) simulations provide a well-understood framework based on Jarzynski's equality to sample from a target probability distribution. By driving a base probability distribution out of equilibrium, observables are computed without the need to thermalize. If the base distribution is characterized by mild autocorrelations, this approach provides a way to mitigate critical slowing down. Out-of-equilibrium evolutions share the same framework of flow-based approaches and they can be naturally combined into a novel architecture called Stochastic Normalizing Flows (SNFs). In this work we present the first implementation of SNFs for $\mathrm{SU}(3)$ lattice gauge theory in 4 dimensions, defined by introducing gauge-equivariant layers between out-of-equilibrium Monte Carlo updates. The core of our analysis is focused on the promising scaling properties of this architecture with the degrees of freedom of the system, which are directly inherited from NE-MCMC. Finally, we discuss how systematic improvements of this approach can realistically lead to a general and yet efficient sampling strategy at fine lattice spacings for observables affected by long autocorrelation times.

14 pages, 12 figures. v2: 14 pages, 13 figures, added comments and improved discussion in section 5, matches published version. v3: fixed figure 7, added missing footnote

References in corpus (28)

• FLAG Review 2021
• Optimal finite-time processes in stochastic thermodynamics
• Critical slowing down and error analysis in lattice QCD simulations
• Optimal protocols for minimal work processes in underdamped stochastic thermodynamics
• The geometry of thermodynamic control
• Optimal Control in Stochastic Thermodynamics
• Efficient Modelling of Trivializing Maps for Lattice ${\varphi }^4$ Theory Using Normalizing Flows: A First Look at Scalability
• Gauge-equivariant flow models for sampling in lattice field theories with pseudofermions
• Fighting topological freezing in the two-dimensional CP${}^{N-1}$ model
• Towards glueball masses of large-$N$ $\mathrm{SU}(N)$ pure-gauge theories without topological freezing
• Flow-based sampling in the lattice Schwinger model at criticality
• Learning Lattice Quantum Field Theories with Equivariant Continuous Flows
• Skipping the Replica Exchange Ladder with Normalizing Flows
• Learning Trivializing Gradient Flows for Lattice Gauge Theories
• Large-$N$ $SU(N)$ Yang-Mills theories with milder topological freezing
• Advances in machine-learning-based sampling motivated by lattice quantum chromodynamics
• Diffusion Models as Stochastic Quantization in Lattice Field Theory
• Entanglement entropy from non-equilibrium Monte Carlo simulations
• Fourier-Flow model generating Feynman paths
• Sampling the lattice Nambu-Goto string using Continuous Normalizing Flows
• Performance of optimal linear-response processes in driven Brownian motion far from equilibrium
• Duality transformations and the entanglement entropy of gauge theories
• Numerical determination of the width and shape of the effective string using Stochastic Normalizing Flows
• On learning higher-order cumulants in diffusion models
• Sampling U(1) gauge theory using a re-trainable conditional flow-based model
• Parallel Tempered Metadynamics: Overcoming potential barriers without surfing or tunneling
• Flow-Based Sampling for Entanglement Entropy and the Machine Learning of Defects
• Exploiting stochastic locality in lattice QCD: hadronic observables and their uncertainties

Cited by in corpus (4)

• Non-Perturbative Trivializing Flows for Lattice Gauge Theories
• Scaling flow-based approaches for topology sampling in $\mathrm{SU}(3)$ gauge theory
• Casimir effect in critical $\mathrm{O}(N)$ models from non-equilibrium Monte Carlo simulations
• Group-Equivariant Diffusion Models for Lattice Field Theory