Symmetries, Safety, and Self-Supervision

Abstract

Collider searches face the challenge of defining a representation of high-dimensional data such that physical symmetries are manifest, the discriminating features are retained, and the choice of representation is new-physics agnostic. We introduce JetCLR to solve the mapping from low-level data to optimized observables though self-supervised contrastive learning. As an example, we construct a data representation for top and QCD jets using a permutation-invariant transformer-encoder network and visualize its symmetry properties. We compare the JetCLR representation with alternative representations using linear classifier tests and find it to work quite well.

References in corpus (12)

• Array Programming with NumPy
• An Introduction to PYTHIA 8.2
• Jet substructure as a new Higgs search channel at the LHC
• Techniques for improved heavy particle searches with jet substructure
• Classification without labels: Learning from mixed samples in high energy physics
• Lorentz Group Equivariant Neural Network for Particle Physics
• Lorentz Boost Networks: Autonomous Physics-Inspired Feature Engineering
• Deep Set Auto Encoders for Anomaly Detection in Particle Physics
• Unsupervised Hadronic SUEP at the LHC
• Shared Data and Algorithms for Deep Learning in Fundamental Physics
• Revealing systematics in phenomenologically viable flux vacua with reinforcement learning
• Particle Convolution for High Energy Physics

Cited by in corpus (28)

• Calomplification -- The Power of Generative Calorimeter Models
• Self-supervised Anomaly Detection for New Physics
• Learning the language of QCD jets with transformers
• OmniJet-$\alpha$: The first cross-task foundation model for particle physics
• Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models
• What's Anomalous in LHC Jets?
• Lorentz group equivariant autoencoders
• A method to challenge symmetries in data with self-supervised learning
• Precision-Machine Learning for the Matrix Element Method
• A Method to Simultaneously Facilitate All Jet Physics Tasks
• Solving Key Challenges in Collider Physics with Foundation Models
• Anomalies, Representations, and Self-Supervision
• Re-Simulation-based Self-Supervised Learning for Pre-Training Foundation Models
• Discovering Sparse Representations of Lie Groups with Machine Learning
• Semi-visible jets, energy-based models, and self-supervision
• SKATR: A Self-Supervised Summary Transformer for SKA
• Unsupervised and lightly supervised learning in particle physics
• Learning the Simplicity of Scattering Amplitudes
• Triggering Dark Showers with Conditional Dual Auto-Encoders
• Top-philic Machine Learning
• Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future colliders
• MACK: Mismodeling Addressed with Contrastive Knowledge
• Contrastive Learning for Robust Representations of Neutrino Data
• Anomaly preserving contrastive neural embeddings for end-to-end model-independent searches at the LHC
• jBOT: Semantic Jet Representation Clustering Emerges from Self-Distillation
• Power Counting Energy Flow Polynomials
• Large Language Models -- the Future of Fundamental Physics?
• Vision Transformers for End-to-End Quark-Gluon Jet Classification from Calorimeter Images