Explainable Equivariant Neural Networks for Particle Physics: PELICAN

Abstract

PELICAN is a novel permutation equivariant and Lorentz invariant or covariant aggregator network designed to overcome common limitations found in architectures applied to particle physics problems. Compared to many approaches that use non-specialized architectures that neglect underlying physics principles and require very large numbers of parameters, PELICAN employs a fundamentally symmetry group-based architecture that demonstrates benefits in terms of reduced complexity, increased interpretability, and raw performance. We present a comprehensive study of the PELICAN algorithm architecture in the context of both tagging (classification) and reconstructing (regression) Lorentz-boosted top quarks, including the difficult task of specifically identifying and measuring the $W$-boson inside the dense environment of the Lorentz-boosted top-quark hadronic final state. We also extend the application of PELICAN to the tasks of identifying quark-initiated vs.~gluon-initiated jets, and a multi-class identification across five separate target categories of jets. When tested on the standard task of Lorentz-boosted top-quark tagging, PELICAN outperforms existing competitors with much lower model complexity and high sample efficiency. On the less common and more complex task of 4-momentum regression, PELICAN also outperforms hand-crafted, non-machine learning algorithms. We discuss the implications of symmetry-restricted architectures for the wider field of machine learning for physics.

arXiv comment52 pages, 34 figures, 12 tables

References in corpus (35)

• XGBoost: A Scalable Tree Boosting System
• The anti-k_t jet clustering algorithm
• FastJet user manual
• DELPHES 3, A modular framework for fast simulation of a generic collider experiment
• Better Jet Clustering Algorithms
• Jet substructure as a new Higgs search channel at the LHC
• Identifying Boosted Objects with N-subjettiness
• Fast inference of deep neural networks in FPGAs for particle physics
• Top-tagging: A Method for Identifying Boosted Hadronic Tops
• ParticleNet: Jet Tagging via Particle Clouds
• Jet Substructure at the Large Hadron Collider: Experimental Review
• Deep learning in color: towards automated quark/gluon jet discrimination
• Strategies to Identify Boosted Tops
• Energy Flow Networks: Deep Sets for Particle Jets
• The Machine Learning Landscape of Top Taggers
• Gaining (Mutual) Information about Quark/Gluon Discrimination
• Deep-learned Top Tagging with a Lorentz Layer
• Quark and Gluon Jet Substructure
• Energy flow polynomials: A complete linear basis for jet substructure
• Finding top quarks with shower deconstruction
• The Lund Jet Plane
• JEDI-net: a jet identification algorithm based on interaction networks
• An Efficient Lorentz Equivariant Graph Neural Network for Jet Tagging
• ABCNet: An attention-based method for particle tagging
• Casimir Meets Poisson: Improved Quark/Gluon Discrimination with Counting Observables
• Precision Jet Substructure from Boosted Event Shapes
• The jet mass distribution after Soft Drop
• Quark-Gluon Tagging: Machine Learning vs Detector
• Point Cloud Transformers applied to Collider Physics
• Lorentz Boost Networks: Autonomous Physics-Inspired Feature Engineering
• Jet Shape Resummation Using Soft-Collinear Effective Theory
• Energy-weighted Message Passing: an infra-red and collinear safe graph neural network algorithm
• BIP: Boost Invariant Polynomials for Efficient Jet Tagging
• Parton Labeling without Matching: Unveiling Emergent Labelling Capabilities in Regression Models
• Lorentz and permutation invariants of particles I

Cited by in corpus (15)

• A Method to Simultaneously Facilitate All Jet Physics Tasks
• A Lorentz-Equivariant Transformer for All of the LHC
• Reconstruction of Unstable Heavy Particles Using Deep Symmetry-Preserving Attention Networks
• Accuracy versus precision in boosted top tagging with the ATLAS detector
• Advancing Tools for Simulation-Based Inference
• Interplay of Traditional Methods and Machine Learning Algorithms for Tagging Boosted Objects
• Foundations of automatic feature extraction at LHC--point clouds and graphs
• Fast Jet Tagging with MLP-Mixers on FPGAs
• Equivariant neural networks for robust $\text{CP}$ observables
• Tagging fully hadronic exotic decays of the vectorlike $\mathbf{B}$ quark using a graph neural network
• Graph theory inspired anomaly detection at the LHC
• The Physics Behind ML-based Quark-Gluon Taggers
• Tagging ultra-boosted jets at FCC-hh using machine learning techniques
• Enhancing generalization in high energy physics using white-box adversarial attacks
• Large Language Models -- the Future of Fundamental Physics?