Preserving gauge invariance in neural networks

Abstract

In these proceedings we present lattice gauge equivariant convolutional neural networks (L-CNNs) which are able to process data from lattice gauge theory simulations while exactly preserving gauge symmetry. We review aspects of the architecture and show how L-CNNs can represent a large class of gauge invariant and equivariant functions on the lattice. We compare the performance of L-CNNs and non-equivariant networks using a non-linear regression problem and demonstrate how gauge invariance is broken for non-equivariant models.

arXiv comment8 pages, 3 figures, proceedings for vConf 2021

References in corpus (10)

• A high-bias, low-variance introduction to Machine Learning for physicists
• Equivariant flow-based sampling for lattice gauge theory
• Gauge Equivariant Convolutional Networks and the Icosahedral CNN
• Sampling using $SU(N)$ gauge equivariant flows
• Regressive and generative neural networks for scalar field theory
• Towards Novel Insights in Lattice Field Theory with Explainable Machine Learning
• Quantum field-theoretic machine learning
• Machine-learning physics from unphysics: Finding deconfinement temperature in lattice Yang-Mills theories from outside the scaling window
• Extending machine learning classification capabilities with histogram reweighting
• Generalization capabilities of translationally equivariant neural networks

Cited by in corpus (1)

• Deep learning lattice gauge theories