Point Cloud Transformers applied to Collider Physics

TL;DR

This paper explores the application of Point Cloud Transformers, a modified Transformer architecture, to collider physics for jet-tagging tasks, demonstrating how advanced sequence models can improve analysis of collision event data.

Contribution

It introduces the use of Point Cloud Transformers for processing unordered particle sets in collider physics, adapting Transformer models to this domain.

Findings

Improved jet-tagging accuracy over traditional methods

Demonstrated effectiveness of Transformer-based models in collider data analysis

Showed potential for enhanced semantic understanding of collision events

Abstract

Methods for processing point cloud information have seen a great success in collider physics applications. One recent breakthrough in machine learning is the usage of Transformer networks to learn semantic relationships between sequences in language processing. In this work, we apply a modified Transformer network called Point Cloud Transformer as a method to incorporate the advantages of the Transformer architecture to an unordered set of particles resulting from collision events. To compare the performance with other strategies, we study jet-tagging applications for highly-boosted particles.