# Stable and Interpretable Jet Physics with IRC-Safe Equivariant Feature Extraction

arXiv: https://arxiv.org/abs/2509.22059 · HTML: https://mlsanity.com/paper/2509.22059/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 92 references that are in the map.
- First posted on arXiv 26 Sep 2025.

## Builds on (most cited first, 26 of 26)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Neural Message Passing for Quantum Chemistry](https://mlsanity.com/paper/1704.01212/) (2017)
- [Particle Transformer for Jet Tagging](https://mlsanity.com/paper/2202.03772/) (2022)
- [Energy Flow Networks: Deep Sets for Particle Jets](https://mlsanity.com/paper/1810.05165/) (2018)
- [ParticleNet: Jet Tagging via Particle Clouds](https://mlsanity.com/paper/1902.08570/) (2019)
- [PyG 2.0: Scalable Learning on Real World Graphs](https://mlsanity.com/paper/2507.16991/) (2025)
- [Machine Learning in High Energy Physics Community White Paper](https://mlsanity.com/paper/1807.02876/) (2018)
- [OmniJet-α: The first cross-task foundation model for particle physics](https://mlsanity.com/paper/2403.05618/) (2024)
- [Jet-Images -- Deep Learning Edition](https://mlsanity.com/paper/1511.05190/) (2015)
- [Machine Learning for Anomaly Detection in Particle Physics](https://mlsanity.com/paper/2312.14190/) (2023)
- [A Lorentz-Equivariant Transformer for All of the LHC](https://mlsanity.com/paper/2411.00446/) (2024)
- [Point Cloud Transformers applied to Collider Physics](https://mlsanity.com/paper/2102.05073/) (2021)
- [Symmetry Group Equivariant Architectures for Physics](https://mlsanity.com/paper/2203.06153/) (2022)
- [QCD-Aware Recursive Neural Networks for Jet Physics](https://mlsanity.com/paper/1702.00748/) (2017)
- [Jet tagging in the Lund plane with graph networks](https://mlsanity.com/paper/2012.08526/) (2020)
- [Is Tokenization Needed for Masked Particle Modelling?](https://mlsanity.com/paper/2409.12589/) (2024)
- [Deep learning in color: towards automated quark/gluon jet discrimination](https://mlsanity.com/paper/1612.01551/) (2016)
- [Explainable AI for ML jet taggers using expert variables and layerwise relevance propagation](https://mlsanity.com/paper/2011.13466/) (2020)
- [Symmetry meets AI](https://mlsanity.com/paper/2103.06115/) (2021)
- [FAIR for AI: An interdisciplinary and international community building perspective](https://mlsanity.com/paper/2210.08973/) (2022)
- [Lorentz Local Canonicalization: How to Make Any Network Lorentz-Equivariant](https://mlsanity.com/paper/2505.20280/) (2025)
- [Interpretable Uncertainty Quantification in AI for HEP](https://mlsanity.com/paper/2208.03284/) (2022)
- [Jet Charge and Machine Learning](https://mlsanity.com/paper/1803.08066/) (2018)
- [Optimal Equivariant Architectures from the Symmetries of Matrix-Element Likelihoods](https://mlsanity.com/paper/2410.18553/) (2024)
- [Artificial Intelligence for the Electron Ion Collider (AI4EIC)](https://mlsanity.com/paper/2307.08593/) (2023)
- [Integrating Physics Inspired Features with Graph Convolution](https://mlsanity.com/paper/2403.11826/) (2024)

## Built on by (most cited first, 2 of 2)

- [Explainable AI for Jet Tagging: A Comparative Study of GNNExplainer, GNNShap, and GradCAM for Jet Tagging in the Lund Jet Plane](https://mlsanity.com/paper/2604.25885/) (2026)
- [Dissecting Jet-Tagger Through Mechanistic Interpretability](https://mlsanity.com/paper/2605.09881/) (2026)

## Abstract (arXiv)

Deep learning has achieved remarkable success in jet classification tasks, yet a key challenge remains: understanding what these models learn and how their features relate to known QCD observables. Improving interpretability is essential for building robust and trustworthy machine learning tools in collider physics. To address this challenge, we investigate graph neural networks for quark-gluon discrimination, systematically incorporating physics-motivated inductive biases. In particular, we design message-passing architectures that enforce infrared and collinear (IRC) safety, as well as E(2) and O(2) equivariance in the rapidity-azimuth plane. Using simulated jet datasets, we compare these networks against unconstrained baselines in terms of classification performance, robustness to soft emissions, and latent representation structures. Our analysis shows that physics-aware networks are more stable across training instances and distribute their latent variance across multiple interpretable directions. By regressing Energy Flow Polynomials onto the leading principal components, we establish a direct correspondence between learned representations and established IRC-safe jet observables. These results demonstrate that embedding symmetry and safety constraints not only improves robustness but also grounds network representations in known QCD structures, providing a principled approach toward interpretable deep learning in collider physics.