The Physics Behind ML-based Quark-Gluon Taggers
arXiv:2507.21214 · doi:10.21468/SciPostPhys.20.3.084
Abstract
Jet taggers provide an ideal testbed for applying explainability techniques to powerful ML tools. For theoretically and experimentally challenging quark-gluon tagging, we first identify the leading latent features that correlate strongly with physics observables, both in a linear and a non-linear approach. Next, we show how Shapley values can assess feature importance, although the standard implementation assumes independent inputs and can lead to distorted attributions in the presence of correlations. Finally, we use symbolic regression to derive compact formulas to approximate the tagger output.
References in corpus (68)
- The anti-k_t jet clustering algorithm
- An Introduction to PYTHIA 8.2
- FastJet user manual
- ParticleNet: Jet Tagging via Particle Clouds
- Energy Correlation Functions for Jet Substructure
- Deep learning in color: towards automated quark/gluon jet discrimination
- Jet-Images -- Deep Learning Edition
- Deep-learning Top Taggers or The End of QCD?
- Energy Flow Networks: Deep Sets for Particle Jets
- Gaining (Mutual) Information about Quark/Gluon Discrimination
- Quark and Gluon Tagging at the LHC
- Les Houches 2015: Physics at TeV Colliders Standard Model Working Group Report
- Interpretable Machine Learning for Science with PySR and SymbolicRegression.jl
- Systematics of quark/gluon tagging
- Identification of heavy, energetic, hadronically decaying particles using machine-learning techniques
- Quark and Gluon Jet Substructure
- Energy flow polynomials: A complete linear basis for jet substructure
- Pulling Out All the Tops with Computer Vision and Deep Learning
- Multivariate discrimination and the Higgs + W/Z search
- Casimir Meets Poisson: Improved Quark/Gluon Discrimination with Counting Observables
- Recursive Neural Networks in Quark/Gluon Tagging
- Search for a Higgs boson in the decay channel H to ZZ(*) to q qbar l-l+ in pp collisions at sqrt(s) = 7 TeV
- Searches for the BSM scenarios at the LHC using decision tree based machine learning algorithms: A comparative study and review of Random Forest, Adaboost, XGboost and LightGBM frameworks
- Quark-Gluon Tagging: Machine Learning vs Detector
- An operational definition of quark and gluon jets
- A Theory of Quark vs. Gluon Discrimination
- Boosted Tagging with Jet Charge and Deep Learning
- The Entropy of a Jet
- Boosted decision trees in the era of new physics: a smuon analysis case study
- Quarks and gluons in the Lund plane
- Modern Machine Learning for LHC Physicists
- Towards Machine Learning Analytics for Jet Substructure
- Interpretable machine learning in Physics
- Explainable Equivariant Neural Networks for Particle Physics: PELICAN
- Quark-Gluon Jet Discrimination Using Convolutional Neural Networks
- A Detailed Study of Interpretability of Deep Neural Network based Top Taggers
- A Lorentz-Equivariant Transformer for All of the LHC
- Quark Gluon Jet Discrimination with Weakly Supervised Learning
- Back to the Formula -- LHC Edition
- Jet Tagging with More-Interaction Particle Transformer
- Discriminating among cosmological models by data-driven methods
- Explainable AI for High Energy Physics
- Systematic Quark/Gluon Identification with Ratios of Likelihoods
- Is infrared-collinear safe information all you need for jet classification?
- Discover the GellMann-Okubo formula with machine learning
- SymbolFit: Automatic Parametric Modeling with Symbolic Regression
- Boosted top tagging and its interpretation using Shapley values
- Unveiling the Secrets of New Physics Through Top Quark Tagging
- A polarization analyzer from Deep Neural Networks
- Secondary Vertex Reconstruction with MaskFormers
- Quark-versus-gluon tagging in CMS Open Data with CWoLa and TopicFlow
- Symbolic Regression for Beyond the Standard Model Physics
- Accelerating Graph-based Tracking Tasks with Symbolic Regression
- Systematic Interpretability and the Likelihood for Boosted Top Quark Identification
- Hierarchical High-Point Energy Flow Network for Jet Tagging
- Quark/Gluon Discrimination and Top Tagging with Dual Attention Transformer
- Performance versus Resilience in Modern Quark-Gluon Tagging
- Symbolic regression for precision LHC physics
- Applicability Evaluation of Selected xAI Methods for Machine Learning Algorithms for Signal Parameters Extraction
- A continuous calibration of the ATLAS flavour-tagging classifiers via optimal transportation maps
- Inferring Interpretable Models of Fragmentation Functions using Symbolic Regression
- What exactly did the Transformer learn from our physics data?
- Application of Machine Learning Based Top Quark and W Jet Tagging to Hadronic Four-Top Final States Induced by SM as well as BSM Processes
- Jet Reconstruction with Mamba Networks in Collider Events
- Proposed measurement of longitudinally polarised vector bosons in and production at Hadron colliders
- IAFormer: Interaction-Aware Transformer network for collider data analysis
- Explainable AI classification for parton density theory
- Product Manifold Machine Learning for Physics