Machine learning for four-dimensional SU(3) lattice gauge theories

2026-04-14

AI Summary

This work addresses the challenges of low sampling efficiency and lattice artifacts in four-dimensional SU(3) lattice gauge theory near the continuum limit by introducing a novel approach that integrates renormalization group (RG) concepts with generative machine learning. The authors construct a gauge-equivariant convolutional neural network to learn an RG-improved effective action and combine stochastic normalizing flows with diffusion processes to enable efficient configuration sampling. This framework yields a machine-learned fixed-point action that closely approximates the continuum limit, substantially suppressing tree-level lattice artifacts. High-precision agreement is demonstrated for key physical observables, including gradient flow scales, static quark potentials, and the deconfinement phase transition, establishing a new paradigm for studying the continuum limit in lattice field theories.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

In this review I summarize how machine learning can be used in lattice gauge theory simulations and what ap\-proaches are currently available to improve the sampling of gauge field configurations, with a focus on applications in four-dimensional SU(3) gauge theories. These include approaches based on generative machine-learning models such as (stochastic) normalizing flows and diffusion processes, and an approach based on renormalization group (RG) transformations, more specifically the machine learning of RG-improved gauge actions using gauge-equivariant convolutional neural networks. In particular, I present scaling results for a machine-learned fixed-point action in four-dimensional SU(3) gauge theory towards the continuum limit. The results include observables based on the classically perfect gradient-flow scales, which are free of tree-level lattice artefacts to all orders, and quantities related to the static potential and the deconfinement transition.

Problem

• lattice gauge theory
• SU(3)
• field configuration sampling
• continuum limit
• lattice artefacts

Innovation

• generative models
• normalizing flows
• diffusion processes
• renormalization group
• gauge-equivariant CNNs