Lorentz Group Equivariant Neural Network for Particle Physics

Abstract

We present a neural network architecture that is fully equivariant with respect to transformations under the Lorentz group, a fundamental symmetry of space and time in physics. The architecture is based on the theory of the finite-dimensional representations of the Lorentz group and the equivariant nonlinearity involves the tensor product. For classification tasks in particle physics, we demonstrate that such an equivariant architecture leads to drastically simpler models that have relatively few learnable parameters and are much more physically interpretable than leading approaches that use CNNs and point cloud approaches. The competitive performance of the network is demonstrated on a public classification dataset [27] for tagging top quark decays given energy-momenta of jet constituents produced in proton-proton collisions.

References in corpus (11)

• Observation of a new particle in the search for the Standard Model Higgs boson with the ATLAS detector at the LHC
• Observation of a new boson at a mass of 125 GeV with the CMS experiment at the LHC
• DELPHES 3, A modular framework for fast simulation of a generic collider experiment
• Deep Convolutional Networks on Graph-Structured Data
• Group Equivariant Convolutional Networks
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• Towards Jetography
• ParticleNet: Jet Tagging via Particle Clouds
• N-body Networks: a Covariant Hierarchical Neural Network Architecture for Learning Atomic Potentials
• Lorentz Boost Networks: Autonomous Physics-Inspired Feature Engineering
• Jet Constituents for Deep Neural Network Based Top Quark Tagging

Cited by in corpus (32)

• Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
• An Efficient Lorentz Equivariant Graph Neural Network for Jet Tagging
• Theoretical Guarantees for Permutation-Equivariant Quantum Neural Networks
• Quasi Anomalous Knowledge: Searching for new physics with embedded knowledge
• Mapping Machine-Learned Physics into a Human-Readable Space
• Equivariant Energy Flow Networks for Jet Tagging
• Symmetries, Safety, and Self-Supervision
• Learning to Simulate High Energy Particle Collisions from Unlabeled Data
• Permutationless Many-Jet Event Reconstruction with Symmetry Preserving Attention Networks
• SymmetryGAN: Symmetry Discovery with Deep Learning
• Shared Data and Algorithms for Deep Learning in Fundamental Physics
• Explainable machine learning of the underlying physics of high-energy particle collisions
• Machine Learning in High Energy Physics: A review of heavy-flavor jet tagging at the LHC
• A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups
• On the Universality of Rotation Equivariant Point Cloud Networks
• A Wigner-Eckart Theorem for Group Equivariant Convolution Kernels
• A Detailed Study of Interpretability of Deep Neural Network based Top Taggers
• Approximately Equivariant Quantum Neural Network for $p4m$ Group Symmetries in Images
• Machine learning and invariant theory
• Coordinate Independent Convolutional Networks -- Isometry and Gauge Equivariant Convolutions on Riemannian Manifolds
• BIP: Boost Invariant Polynomials for Efficient Jet Tagging
• Quantum algorithms for group convolution, cross-correlation, and equivariant transformations
• Attention to the strengths of physical interactions: Transformer and graph-based event classification for particle physics experiments
• ${\mathbb{Z}}_2\times {\mathbb{Z}}_2$ Equivariant Quantum Neural Networks: Benchmarking against Classical Neural Networks
• Lorentz and permutation invariants of particles III: constraining non-standard sources of parity violation
• Invariant polynomials and machine learning
• Beyond permutation equivariance in graph networks
• Capacity of Group-invariant Linear Readouts from Equivariant Representations: How Many Objects can be Linearly Classified Under All Possible Views?
• Autoequivariant Network Search via Group Decomposition
• A FAIR and AI-ready Higgs boson decay dataset
• Computing equivariant matrices on homogeneous spaces for Geometric Deep Learning and Automorphic Lie Algebras
• Large Language Models -- the Future of Fundamental Physics?