Particle transformers for identifying Lorentz-boosted Higgs bosons decaying to a pair of W bosons
8 affiliations
- 1Yerevan Physics Institutea
- 2Yerevan State University
- 3Institut für Hochenergiephysik
- 4Technische Universität Wien
- 5Universiteit Antwerpen
- 6Vrije Universiteit Brussel
- 7Ghent University
- 8Université Libre de Bruxelles
Abstract
A bstract : A novel deep neural network classifier, a “particle transformer” (P ar T), is introduced for the identification of highly Lorentz-boosted resonances reconstructed as single, multipronged jets in measurements and searches performed by the CMS Collaboration at the CERN LHC. Based on a self-attention mechanism that allows the model to weigh the importance of different particles, P ar T is trained on a wide variety of topologies, notably demonstrating strong performance for the first time on jets originating from boosted Higgs boson decays to W bosons. The P ar T algorithm achieves a tagging efficiency of more than 50% for such jets at a background efficiency of 1%, while maintaining decorrelation from the jet mass. A calibration is performed in proton-proton collision data collected by CMS at a center-of-mass energy of 13 TeV, with a data set corresponding to a total luminosity of 138 fb − 1 . Data-to-simulation selection efficiency scale factors are measured to be in the 0.9–1.0 range, with relative uncertainties between 7 and 23%. The tagging capability of P ar T enhances the sensitivity of standard model measurements and searches for beyond-the-standard-model resonances decaying to hadronic diboson systems.
Related
- Search for pair-produced vector-like top quarks decaying into Lorentz-boosted top quarks and scalar bosons in proton-proton collisions at √s = 13 TeV
- Lorentz-Equivariant Geometric Algebra Transformers for Top-Quark Event Reconstruction
- Evolution of inertial particle clustering in decaying turbulence
- End-to-end Particle Reconstruction with Transformers in High-Granularity Calorimeters
- Conditional Capacity and Routing in Mixture-of-Experts Particle Transformers
BibTeX
@article{collaboration2026particle,
title = {{Particle transformers for identifying Lorentz-boosted Higgs bosons decaying to a pair of W bosons}},
author = {The CMS collaboration and A. Hayrapetyan and V. Makarenko and A. Tumasyan and W. Adam and L. Benato and T. Bergauer and M. Dragicevic and C. Giordano and P. S. Hussain and M. Jeitler and N. Krammer and A. Li and D. Liko and M. Matthewman and J. Schieck and R. Schöfbeck and M. Shooshtari and M. Sonawane and W. Waltenberger and C. E. Wulz and T. Janssen and H. Kwon and D. Ocampo Henao and T. Van Laer and P. Van Mechelen and J. Bierkens and N. Breugelmans and J. D’Hondt and S. Dansana and A. De Moor and M. Delcourt and F. Heyen and Y. Hong and P. Kashko and S. Lowette and I. Makarenko and D. Müller and J. Song and S. Tavernier and M. Tytgat and G. P. Van Onsem and S. Van Putte and D. Vannerom and B. Bilin and B. Clerbaux and A. K. Das and I. De Bruyn and G. De Lentdecker and H. Evard and L. Favart and P. Gianneios and A. Khalilzadeh and F. A. Khan and A. Malara and M. A. Shahzad and A. Sharma and L. Thomas and M. Vanden Bemden and C. Vander Velde},
journal = {Journal of High Energy Physics},
year = {2026},
month = oct,
doi = {10.1007/jhep10(2026)167},
url = {https://link.springer.com/article/10.1007/JHEP10(2026)167},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Particle transformers for identifying Lorentz-boosted Higgs bosons decaying to a pair of W bosons
AU - The CMS collaboration
AU - A. Hayrapetyan
AU - V. Makarenko
AU - A. Tumasyan
AU - W. Adam
AU - L. Benato
AU - T. Bergauer
AU - M. Dragicevic
AU - C. Giordano
AU - P. S. Hussain
AU - M. Jeitler
AU - N. Krammer
AU - A. Li
AU - D. Liko
AU - M. Matthewman
AU - J. Schieck
AU - R. Schöfbeck
AU - M. Shooshtari
AU - M. Sonawane
AU - W. Waltenberger
AU - C. E. Wulz
AU - T. Janssen
AU - H. Kwon
AU - D. Ocampo Henao
AU - T. Van Laer
AU - P. Van Mechelen
AU - J. Bierkens
AU - N. Breugelmans
AU - J. D’Hondt
AU - S. Dansana
AU - A. De Moor
AU - M. Delcourt
AU - F. Heyen
AU - Y. Hong
AU - P. Kashko
AU - S. Lowette
AU - I. Makarenko
AU - D. Müller
AU - J. Song
AU - S. Tavernier
AU - M. Tytgat
AU - G. P. Van Onsem
AU - S. Van Putte
AU - D. Vannerom
AU - B. Bilin
AU - B. Clerbaux
AU - A. K. Das
AU - I. De Bruyn
AU - G. De Lentdecker
AU - H. Evard
AU - L. Favart
AU - P. Gianneios
AU - A. Khalilzadeh
AU - F. A. Khan
AU - A. Malara
AU - M. A. Shahzad
AU - A. Sharma
AU - L. Thomas
AU - M. Vanden Bemden
AU - C. Vander Velde
PY - 2026
DA - 2026/10/15
JO - Journal of High Energy Physics
DO - 10.1007/jhep10(2026)167
UR - https://link.springer.com/article/10.1007/JHEP10(2026)167
AB - A bstract : A novel deep neural network classifier, a “particle transformer” (P ar T), is introduced for the identification of highly Lorentz-boosted resonances reconstructed as single, multipronged jets in measurements and searches performed by the CMS Collaboration at the CERN LHC. Based on a self-attention mechanism that allows the model to weigh the importance of different particles, P ar T is trained on a wide variety of topologies, notably demonstrating strong performance for the first time on jets originating from boosted Higgs boson decays to W bosons. The P ar T algorithm achieves a tagging efficiency of more than 50% for such jets at a background efficiency of 1%, while maintaining decorrelation from the jet mass. A calibration is performed in proton-proton collision data collected by CMS at a center-of-mass energy of 13 TeV, with a data set corresponding to a total luminosity of 138 fb − 1 . Data-to-simulation selection efficiency scale factors are measured to be in the 0.9–1.0 range, with relative uncertainties between 7 and 23%. The tagging capability of P ar T enhances the sensitivity of standard model measurements and searches for beyond-the-standard-model resonances decaying to hadronic diboson systems.
SN - 1029-8479
ER -