Machine Learning of Thermodynamic Observables in the Presence of Mode Collapse

Abstract

Estimating the free energy, as well as other thermodynamic observables, is a key task in lattice field theories. Recently, it has been pointed out that deep generative models can be used in this context [1]. Crucially, these models allow for the direct estimation of the free energy at a given point in parameter space. This is in contrast to existing methods based on Markov chains which generically require integration through parameter space. In this contribution, we will review this novel machine-learning-based estimation method. We will in detail discuss the issue of mode collapse and outline mitigation techniques which are particularly suited for applications at finite temperature.

10 pages, 2 figures, Proceedings of the 38th International Symposium on Lattice Field Theory, 26th-30th July 2021, Zoom/Gather@Massachusetts Institute of Technology

References in corpus (6)

• Normalizing flows for atomic solids
• Quantum field-theoretic machine learning
• Flow-based sampling for multimodal and extended-mode distributions in lattice field theory
• Towards reduction of autocorrelation in HMC by machine learning
• Generalization capabilities of translationally equivariant neural networks
• Comment on "Solving Statistical Mechanics Using VANs": Introducing saVANt - VANs Enhanced by Importance and MCMC Sampling

Cited by in corpus (6)

• Learning Lattice Quantum Field Theories with Equivariant Continuous Flows
• Learning Trivializing Gradient Flows for Lattice Gauge Theories
• Flow-based density of states for complex actions
• Parallel Tempered Metadynamics: Overcoming potential barriers without surfing or tunneling
• AdvNF: Reducing Mode Collapse in Conditional Normalising Flows using Adversarial Learning
• Applications of flow models to the generation of correlated lattice QCD ensembles