Symmetries, safety, and self-supervision

• Verfasst von: Dillon, Barry M. [VerfasserIn]
• Verfasst von: Kasieczka, Gregor [VerfasserIn]
• Verfasst von: Olischläger, Hans [VerfasserIn]
• Verfasst von: Plehn, Tilman [VerfasserIn]
• Verfasst von: Sorrenson, Peter [VerfasserIn]
• Verfasst von: Vogel, Lorenz [VerfasserIn]
• Titel: Symmetries, safety, and self-supervision
• Verf.angabe: Barry Dillon, Gregor Kasieczka, Hans Olischlager, Tilman Plehn, Peter Sorrenson and Lorenz Vogel
• E-Jahr: 2022
• Jahr: 9 June 2022
• Umfang: 19 S.
• Fussnoten: Gesehen am 14.07.2022
• Titel Quelle: Enthalten in: SciPost physics
• Ort Quelle: Amsterdam : SciPost Foundation, 2016
• Jahr Quelle: 2022
• Band/Heft Quelle: 12(2022), 6, Artikel-ID 188, Seite 1-19
• ISSN Quelle: 2542-4653
• DOI: doi:10.21468/SciPostPhys.12.6.188
• Datenträger: Online-Ressource
• Sprache: eng
• K10plus-PPN: 1810190967

Abstract

Collider searches face the challenge of defining a representation of high-dimensional data such that physical symmetries are manifest, the discriminating features are retained, and the choice of representation is new-physics agnostic. We introduce JetCLR to solve the mapping from low-level data to optimized observables though self-supervised contrastive learning. As an example, we construct a data representation for top and QCD jets using a permutation-invariant transformer-encoder network and visualize its symmetry properties. We compare the JetCLR representation with alternative representations using linear classifier tests and find it to work quite well.