Particle Multi-Axis Transformer for Jet Tagging

TL;DR

The paper introduces Particle Multi-Axis Transformer (ParMAT), a novel deep learning architecture for jet tagging in high energy physics, combining local and global interactions to improve accuracy and scalability.

Contribution

It proposes a new transformer architecture, ParMAT, with integrated local and global spatial interactions, enhancing jet tagging performance over existing models.

Findings

ParMAT outperforms ParT and ParticleNet in accuracy.

The model demonstrates robustness across various input lengths.

ParMAT scales effectively to large datasets.

Abstract

Jet tagging is an essential categorization problem in high energy physics. In recent times, Deep Learning has not only risen to the challenge of jet tagging but also significantly improved its performance. In this article, we proposed an idea of a new architecture, Particle Multi-Axis transformer (ParMAT) which is a modified version of Particle transformer (ParT). ParMAT contains local and global spatial interactions within a single unit which improves its ability to handle various input lengths. We trained our model on JETCLASS, a publicly available large dataset that contains 100M jets of 10 different classes of particles. By integrating a parallel attention mechanism and pairwise interactions of particles in the attention mechanism, ParMAT achieves robustness and higher accuracy over the ParT and ParticleNet. The scalability of the model to huge datasets and its ability to automatically extract essential features demonstrate its potential for enhancing jet tagging.