A Lorentz-Equivariant Transformer for All of the LHC

A Lorentz-Equivariant Transformer for All of the LHC

We show that the Lorentz-Equivariant Geometric Algebra Transformer (L-GATr) yields state-of-the-art performance for a wide range of machine learning tasks at the Large Hadron Collider. L-GATr represents data in a geometric algebra over space-time and is equivariant under Lorentz transformations. The...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Brehmer, Johann, Bresó, Víctor, de Haan, Pim, Plehn, Tilman, Qu, Huilin, Spinner, Jonas, Thaler, Jesse
Format: Artikel
Sprache:eng
Schlagworte:
Computer Science - Learning
Physics - High Energy Physics - Experiment
Physics - High Energy Physics - Phenomenology
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Brehmer, Johann
Bresó, Víctor
de Haan, Pim
Plehn, Tilman
Qu, Huilin
Spinner, Jonas
Thaler, Jesse
description We show that the Lorentz-Equivariant Geometric Algebra Transformer (L-GATr) yields state-of-the-art performance for a wide range of machine learning tasks at the Large Hadron Collider. L-GATr represents data in a geometric algebra over space-time and is equivariant under Lorentz transformations. The underlying architecture is a versatile and scalable transformer, which is able to break symmetries if needed. We demonstrate the power of L-GATr for amplitude regression and jet classification, and then benchmark it as the first Lorentz-equivariant generative network. For all three LHC tasks, we find significant improvements over previous architectures.
doi_str_mv 10.48550/arxiv.2411.00446
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2411_00446</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2411_00446</sourcerecordid><originalsourceid>FETCH-arxiv_primary_2411_004463</originalsourceid><addsrcrecordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjE01DMwMDEx42QwcVTwyS9KzSup0nUtLM0sSyzKTMwrUQgpSswrTssvyk0tUgBSCo45OQr5aQolGakKPh7OPAysaYk5xam8UJqbQd7NNcTZQxdsfnxBUWZuYlFlPMieeLA9xoRVAADXljFP</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>A Lorentz-Equivariant Transformer for All of the LHC</title><source>arXiv.org</source><creator>Brehmer, Johann ; Bresó, Víctor ; de Haan, Pim ; Plehn, Tilman ; Qu, Huilin ; Spinner, Jonas ; Thaler, Jesse</creator><creatorcontrib>Brehmer, Johann ; Bresó, Víctor ; de Haan, Pim ; Plehn, Tilman ; Qu, Huilin ; Spinner, Jonas ; Thaler, Jesse</creatorcontrib><description>We show that the Lorentz-Equivariant Geometric Algebra Transformer (L-GATr) yields state-of-the-art performance for a wide range of machine learning tasks at the Large Hadron Collider. L-GATr represents data in a geometric algebra over space-time and is equivariant under Lorentz transformations. The underlying architecture is a versatile and scalable transformer, which is able to break symmetries if needed. We demonstrate the power of L-GATr for amplitude regression and jet classification, and then benchmark it as the first Lorentz-equivariant generative network. For all three LHC tasks, we find significant improvements over previous architectures.</description><identifier>DOI: 10.48550/arxiv.2411.00446</identifier><language>eng</language><subject>Computer Science - Learning ; Physics - High Energy Physics - Experiment ; Physics - High Energy Physics - Phenomenology</subject><creationdate>2024-11</creationdate><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2411.00446$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2411.00446$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Brehmer, Johann</creatorcontrib><creatorcontrib>Bresó, Víctor</creatorcontrib><creatorcontrib>de Haan, Pim</creatorcontrib><creatorcontrib>Plehn, Tilman</creatorcontrib><creatorcontrib>Qu, Huilin</creatorcontrib><creatorcontrib>Spinner, Jonas</creatorcontrib><creatorcontrib>Thaler, Jesse</creatorcontrib><title>A Lorentz-Equivariant Transformer for All of the LHC</title><description>We show that the Lorentz-Equivariant Geometric Algebra Transformer (L-GATr) yields state-of-the-art performance for a wide range of machine learning tasks at the Large Hadron Collider. L-GATr represents data in a geometric algebra over space-time and is equivariant under Lorentz transformations. The underlying architecture is a versatile and scalable transformer, which is able to break symmetries if needed. We demonstrate the power of L-GATr for amplitude regression and jet classification, and then benchmark it as the first Lorentz-equivariant generative network. For all three LHC tasks, we find significant improvements over previous architectures.</description><subject>Computer Science - Learning</subject><subject>Physics - High Energy Physics - Experiment</subject><subject>Physics - High Energy Physics - Phenomenology</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjE01DMwMDEx42QwcVTwyS9KzSup0nUtLM0sSyzKTMwrUQgpSswrTssvyk0tUgBSCo45OQr5aQolGakKPh7OPAysaYk5xam8UJqbQd7NNcTZQxdsfnxBUWZuYlFlPMieeLA9xoRVAADXljFP</recordid><startdate>20241101</startdate><enddate>20241101</enddate><creator>Brehmer, Johann</creator><creator>Bresó, Víctor</creator><creator>de Haan, Pim</creator><creator>Plehn, Tilman</creator><creator>Qu, Huilin</creator><creator>Spinner, Jonas</creator><creator>Thaler, Jesse</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20241101</creationdate><title>A Lorentz-Equivariant Transformer for All of the LHC</title><author>Brehmer, Johann ; Bresó, Víctor ; de Haan, Pim ; Plehn, Tilman ; Qu, Huilin ; Spinner, Jonas ; Thaler, Jesse</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-arxiv_primary_2411_004463</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Learning</topic><topic>Physics - High Energy Physics - Experiment</topic><topic>Physics - High Energy Physics - Phenomenology</topic><toplevel>online_resources</toplevel><creatorcontrib>Brehmer, Johann</creatorcontrib><creatorcontrib>Bresó, Víctor</creatorcontrib><creatorcontrib>de Haan, Pim</creatorcontrib><creatorcontrib>Plehn, Tilman</creatorcontrib><creatorcontrib>Qu, Huilin</creatorcontrib><creatorcontrib>Spinner, Jonas</creatorcontrib><creatorcontrib>Thaler, Jesse</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Brehmer, Johann</au><au>Bresó, Víctor</au><au>de Haan, Pim</au><au>Plehn, Tilman</au><au>Qu, Huilin</au><au>Spinner, Jonas</au><au>Thaler, Jesse</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>A Lorentz-Equivariant Transformer for All of the LHC</atitle><date>2024-11-01</date><risdate>2024</risdate><abstract>We show that the Lorentz-Equivariant Geometric Algebra Transformer (L-GATr) yields state-of-the-art performance for a wide range of machine learning tasks at the Large Hadron Collider. L-GATr represents data in a geometric algebra over space-time and is equivariant under Lorentz transformations. The underlying architecture is a versatile and scalable transformer, which is able to break symmetries if needed. We demonstrate the power of L-GATr for amplitude regression and jet classification, and then benchmark it as the first Lorentz-equivariant generative network. For all three LHC tasks, we find significant improvements over previous architectures.</abstract><doi>10.48550/arxiv.2411.00446</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2411.00446
ispartof
issn
language eng
recordid cdi_arxiv_primary_2411_00446
source arXiv.org
subjects Computer Science - Learning
Physics - High Energy Physics - Experiment
Physics - High Energy Physics - Phenomenology
title A Lorentz-Equivariant Transformer for All of the LHC
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-27T02%3A15%3A44IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=A%20Lorentz-Equivariant%20Transformer%20for%20All%20of%20the%20LHC&rft.au=Brehmer,%20Johann&rft.date=2024-11-01&rft_id=info:doi/10.48550/arxiv.2411.00446&rft_dat=%3Carxiv_GOX%3E2411_00446%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true