Transforming jet flavour tagging at ATLAS



Aad, G, Aakvaag, E, Abbott, B, Abdelhameed, S, Abeling, K, Abicht, NJ, Abidi, SH, Aboelela, M, Aboulhorma, A, Abramowicz, H
et al (show 2932 more authors) (2026) Transforming jet flavour tagging at ATLAS NATURE COMMUNICATIONS, 17 (1). 541-. ISSN 2041-1723, 2041-1723

Access the full-text of this item by clicking on the Open Access link.

Abstract

Jet flavour tagging enables the identification of jets originating from heavy-flavour quarks in proton–proton collisions at the Large Hadron Collider, playing a critical role in its physics programmes. This paper presents GN2, a transformer-based flavour tagging algorithm deployed by the ATLAS Collaboration that represents a different methodology compared to previous approaches. Designed to classify jets based on the flavour of their constituent particles, GN2 processes low-level tracking information in an end-to-end architecture and incorporates physics-informed auxiliary training objectives to enhance both interpretability and performance. Its performance is validated in both simulation and collision data. The measured c-jet (light-jet) rejection in data is improved by a factor of 3.5 (1.8) for a 70% b-jet tagging efficiency, compared to the previous algorithm. GN2 provides substantial benefits for physics analyses involving heavy-flavour jets, such as measurements of Higgs boson pair production and the couplings of bottom and charm quarks to the Higgs boson, and demonstrates the impact of advanced machine learning methods in experimental particle physics.

Item Type: Article
Uncontrolled Keywords: ATLAS Collaboration
Divisions: Faculty of Science & Engineering
Faculty of Science & Engineering > School of Physical Sciences
Faculty of Science & Engineering > School of Physical Sciences > Physics
Depositing User: Symplectic Admin
Date Deposited: 02 Mar 2026 11:42
Last Modified: 16 Jun 2026 05:44
DOI: 10.1038/s41467-025-65059-6
Open Access URL: https://www.nature.com/articles/s41467-025-65059-6
Related Websites:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3197300
Disclaimer: The University of Liverpool is not responsible for content contained on other websites from links within repository metadata. Please contact us if you notice anything that appears incorrect or inappropriate.