Reducing Autocorrelation Times in Lattice Simulations with Generative Adversarial Networks

Abstract

Short autocorrelation times are essential for a reliable error assessment in Monte Carlo simulations of lattice systems. In many interesting scenarios, the decay of autocorrelations in the Markov chain is prohibitively slow. Generative samplers can provide statistically independent field configurations, thereby potentially ameliorating these issues. In this work, the applicability of neural samplers to this problem is investigated. Specifically, we work with a generative adversarial network (GAN). We propose to address difficulties regarding its statistical exactness through the implementation of an overrelaxation step, by searching the latent space of the trained generator network. This procedure can be incorporated into a standard Monte Carlo algorithm, which then permits a sensible assessment of ergodicity and balance based on consistency checks. Numerical results for real, scalar ${\varphi }^4$-theory in two dimensions are presented. We achieve a significant reduction of autocorrelations while accurately reproducing the correct statistics. We discuss possible improvements to the approach as well as potential solutions to persisting issues.

9 pages, 9 figures

References in corpus (16)

• Learning phase transitions by confusion
• Discovering Phases, Phase Transitions and Crossovers through Unsupervised Machine Learning: A critical examination
• Restricted-Boltzmann-Machine Learning for Solving Strongly Correlated Quantum Systems
• Self-Learning Monte Carlo Method
• Accelerate Monte Carlo Simulations with Restricted Boltzmann Machines
• Solving Statistical Mechanics Using Variational Autoregressive Networks
• Flow-based generative models for Markov chain Monte Carlo in lattice field theory
• Equivariant flow-based sampling for lattice gauge theory
• Machine Learning of Explicit Order Parameters: From the Ising Model to SU(2) Lattice Gauge Theory
• Machine Learning for Quantum Matter
• Asymptotically unbiased estimation of physical observables with neural samplers
• Regressive and generative neural networks for scalar field theory
• Thermodynamics and Feature Extraction by Machine Learning
• Towards Novel Insights in Lattice Field Theory with Explainable Machine Learning
• Restricted Boltzmann Machines for the Long Range Ising Models
• Self-learning Monte Carlo method with Behler-Parrinello neural networks

Cited by in corpus (28)

• Machine Learning in Nuclear Physics
• Machine Learning for Quantum Matter
• Estimation of Thermodynamic Observables in Lattice Field Theories with Deep Generative Models
• Sampling using $SU(N)$ gauge equivariant flows
• CaloFlow: Fast and Accurate Generation of Calorimeter Showers with Normalizing Flows
• Exploring QCD matter in extreme conditions with Machine Learning
• Spectral Reconstruction with Deep Neural Networks
• Lattice gauge equivariant convolutional neural networks
• Efficient Modelling of Trivializing Maps for Lattice ${\varphi }^4$ Theory Using Normalizing Flows: A First Look at Scalability
• Flow-based sampling for fermionic lattice field theories
• Gauge-equivariant flow models for sampling in lattice field theories with pseudofermions
• Stochastic normalizing flows as non-equilibrium transformations
• Physics-Driven Learning for Inverse Problems in Quantum Chromodynamics
• Learning Lattice Quantum Field Theories with Equivariant Continuous Flows
• Machine-learning physics from unphysics: Finding deconfinement temperature in lattice Yang-Mills theories from outside the scaling window
• Advances in machine-learning-based sampling motivated by lattice quantum chromodynamics
• Generative models for sampling and phase transition indication in spin systems
• Diffusion Models as Stochastic Quantization in Lattice Field Theory
• Fourier-Flow model generating Feynman paths
• Continuous-mixture Autoregressive Networks for efficient variational calculation of many-body systems
• Parameter Estimation using Neural Networks in the Presence of Detector Effects
• Optical lattice experiments at unobserved conditions and scales through generative adversarial deep learning
• Accelerating equilibrium spin-glass simulations using quantum annealers via generative deep learning
• Generative learning for the problem of critical slowing down in lattice Gross Neveu model
• Sampling U(1) gauge theory using a re-trainable conditional flow-based model
• Generative Neural Samplers for the Quantum Heisenberg Chain
• Gauge covariant neural network for quarks and gluons
• Group-Equivariant Diffusion Models for Lattice Field Theory