14–18 Sept 2026
Europe/Vienna timezone

Group equivariance in infrared and collinear safe graph neural networks for jet classification

18 Sept 2026, 11:10
20m

Speaker

Vishal Singh Ngairangbam

Description

Improving interpretability is essential for building robust
and trustworthy machine learning tools in collider physics. To address this challenge, we systematically investigate equivariant and IRC-safe graph neural networks for jet classification. Using simulated jet datasets, we compare IRC-safe architectures with inbuilt E(2) and O(2) equivariance in the rapidity-azimuth plane against IRC-safe and -unsafe baselines in terms of classification performance, robustness to soft emissions, and latent representation structures. Our analysis shows that IRC-safe and symmetry-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.

Authors

Partha Konar (Physical Research Laboratory, Ahmedabad, Gujarat-380 009, INDIA) Vishal Singh Ngairangbam Michael Spannowsky (Karlsruhe Institute of Technology) Deepanshu Srivastava (Physical Research Laboratory)

Presentation materials

There are no materials yet.