Classifying hadronic objects in ATLAS with ML/AI algorithms

TL;DR

This paper reviews recent AI-driven methods for classifying hadronic objects in ATLAS, highlighting advances in neural network architectures and their effectiveness in particle physics analyses.

Contribution

It introduces new constituent-based AI architectures like GNNs and transformers for jet classification, demonstrating improved performance in simulated and real data.

Findings

Graph neural networks outperform traditional methods in jet tagging

Transformer-based models show promising results in identifying heavy hadronic objects

Data-driven optimization enhances model robustness and accuracy

Abstract

The identification of hadronic final states plays a crucial role in the physics programme of the ATLAS Experiment at the CERN LHC. Sophisticated artificial intelligence (AI) algorithms are employed to classify jets according to their origin, distinguishing between quark- and gluon-initiated jets, and identifying hadronically decaying heavy objects such as W bosons and top quarks. This contribution summarises recent developments in constituent-based tagging architectures, including graph neural networks (GNNs) and transformer-based approaches, their performance in simulated and real data, and future perspectives towards data-driven optimisation and model-independent tagging strategies.