<p>Hadronization models used in event generators are physics-inspired functions with many tunable parameters. Since we do not understand hadronization from first principles, there have been multiple proposals to improve the accuracy of hadronization models by utilizing more flexible parametrizations based on neural networks. These recent proposals have focused on the kinematic properties of hadrons, but a full model must also include particle flavor. In this paper, we show how to build a deep learning-based hadronization model that includes both kinematic (continuous) and flavor (discrete) degrees of freedom. Our approach is based on generative adversarial networks and we show the performance within the context of the cluster hadronization model within the <span class="sc">h</span>erwig event generator.</p>

United States

Jay Chan

Xiangyang Ju

Poland

Adam Kania

Benjamin Nachman

Vishnu Sangli

Andrzej Siodmok

Integrating particle flavor into deep learning models for hadronization

Hadronization is a critical step in the simulation of high-energy particle and nuclear physics experiments. As there is no first principles understanding of this process, physically-inspired hadronization models have a large number of parameters that are fit to data. Deep generative models are a natural replacement for classical techniques, since they are more flexible and may be able to improve the overall precision. Proof of principle studies have shown how to use neural networks to emulate specific hadronization when trained using the inputs and outputs of classical methods. However, these approaches will not work with data, where we do not have a matching between observed hadrons and partons. In this paper, we develop a protocol for fitting a deep generative hadronization model in a realistic setting, where we only have access to a set of hadrons in data. Our approach uses a variation of a Generative Adversarial Network with a permutation invariant discriminator. We find that this setup is able to match the hadronization model in Herwig with multiple sets of parameters. This work represents a significant step forward in a longer term program to develop, train, and integrate machine learning-based hadronization models into parton shower Monte Carlo programs.

Andrzej Siódmok

Fitting a deep generative hadronization model

Parton labeling methods are widely used when reconstructing collider events with top quarks or other massive particles. State-of-the-art techniques are based on machine learning and require training data with events that have been matched using simulations with truth information. In nature, there is no unique matching between partons and final state objects due to the properties of the strong force and due to acceptance effects. We propose a new approach to parton labeling that circumvents these challenges by recycling regression models. The final state objects that are most relevant for a regression model to predict the properties of a particular top quark are assigned to said parent particle without having any parton-matched training data. This approach is demonstrated using simulated events with top quarks and outperforms the widely-used  $$\chi ^2$$  <math> <msup> <mi>χ</mi> <mn>2</mn> </msup> </math>   method.

Shikai Qiu

Shuo Han

Haichen Wang

Parton labeling without matching: unveiling emergent labelling capabilities in regression models

Precise reconstruction of top quark properties is a challenging task at the Large Hadron Collider due to combinatorial backgrounds and missing information. We introduce a physics-informed neural network architecture called the covariant particle transformer (<math><mrow><mi>C</mi><mi>P</mi><mi>T</mi></mrow></math>) for directly predicting the top quark kinematic properties from reconstructed final state objects. This approach is permutation invariant and partially Lorentz covariant and can account for a variable number of input objects. In contrast to previous machine learning-based reconstruction methods, <math><mrow><mi>C</mi><mi>P</mi><mi>T</mi></mrow></math> is able to predict top quark four-momenta regardless of the jet multiplicity in the event. Using simulations, we show that the <math><mrow><mi>C</mi><mi>P</mi><mi>T</mi></mrow></math> performs favorably compared with other machine learning top quark reconstruction approaches.

Holistic approach to predicting top quark kinematic properties with the covariant particle transformer

Hadronization is a complex quantum process whereby quarks and gluons become hadrons. The widely used models of hadronization in event generators are based on physically inspired phenomenological models with many free parameters. We propose an alternative approach whereby neural networks are used instead. Deep generative models are highly flexible, differentiable, and compatible with graphical processing units. We make the first step towards a data-driven machine learning-based hadronization model. In that step, we replace a component of the hadronization model within the Herwig event generator (cluster model) with <math><mi>HADML</mi></math>, a computer code implementing a generative adversarial network. We show that a <math><mi>HADML</mi></math> is capable of reproducing the kinematic properties of cluster decays. Furthermore, we integrate it into <math><mi>Herwig</mi></math> to generate entire events that can be compared with the output of the public <math><mi>Herwig</mi></math> simulator as well as with <math><msup><mi>e</mi><mo>+</mo></msup><msup><mi>e</mi><mo>−</mo></msup></math> data.

Aishik Ghosh

Towards a deep learning model for hadronization

The Exa.TrkX project has applied geometric learning concepts such as metric learning and graph neural networks to HEP particle tracking. Exa.TrkX’s tracking pipeline groups detector measurements to form track candidates and filters them. The pipeline, originally developed using the TrackML dataset (a simulation of an LHC-inspired tracking detector), has been demonstrated on other detectors, including DUNE Liquid Argon TPC and CMS High-Granularity Calorimeter. This paper documents new developments needed to study the physics and computing performance of the Exa.TrkX pipeline on the full TrackML dataset, a first step towards validating the pipeline using ATLAS and CMS data. The pipeline achieves tracking efficiency and purity similar to production tracking algorithms. Crucially for future HEP applications, the pipeline benefits significantly from GPU acceleration, and its computational requirements scale close to linearly with the number of particles in the event.

Daniel Murnane

Paolo Calafiura

Nicholas Choma

Sean Conlon

Steven Farrell

Yaoyuan Xu

Maria Spiropulu

Jean-Roch Vlimant

Adam Aurisano

Jeremy Hewes

Giuseppe Cerati

Lindsey Gray

Thomas Klijnsma

Jim Kowalkowski

Markus Atkinson

Mark Neubauer

Gage DeZoort

Savannah Thais

Aditi Chauhan

Alex Schuy

Shih-Chieh Hsu

Alex Ballow

Alina Lazar

Performance of a geometric deep learning pipeline for HL-LHC particle tracking

Jet clustering is traditionally an unsupervised learning task because there is no unique way to associate hadronic final states with the quark and gluon degrees of freedom that generated them. However, for uncolored particles like <math><mi>W</mi></math>, <math><mi>Z</mi></math>, and Higgs bosons, it is possible to approximately (though not exactly) associate final state hadrons to their ancestor. By labeling simulated final state hadrons as descending from an uncolored particle, it is possible to train a supervised learning method to create boson jets. Such a method would operate on individual particles and identify connections between particles originating from the same uncolored particle. Graph neural networks are well-suited for this purpose as they can act on unordered sets and naturally create strong connections between particles with the same label. These networks are used to train a supervised jet clustering algorithm. The kinematic properties of these graph jets better match the properties of simulated Lorentz-boosted <math><mi>W</mi></math> bosons. Furthermore, the graph jets contain more information for discriminating <math><mi>W</mi></math> jets from generic quark jets. This work marks the beginning of a new exploration in jet physics to use machine learning to optimize the construction of jets and not only the observables computed from jet constituents.

Supervised jet clustering with graph neural networks for Lorentz boosted bosons

In this study, we investigate the invariant-mass distribution of top-quark pairs near the 2mt threshold, which strongly influences the determination of the top-quark mass mt. Higher-order non-relativistic corrections lead to large contributions, which are not included in the state-of-the-art theoretical predictions. A factorization formula is derived to resum such corrections to all orders in the strong-coupling, and necessary ingredients are calculated to perform the resummation at next-to-leading power. We combine the resummation with fixed-order results and present phenomenologically relevant numerical results. The resummation effect significantly increases the differential cross-section in the threshold region and makes the theoretical prediction more compatible with experimental data. We estimate that using our prediction in the determination of mt will lead to a value closer to the direct measurement result.
				

China

United Kingdom

Wan-Li Ju

Guoxing Wang

Germany

Xing Wang

Switzerland

Xiaofeng Xu

Yongqi Xu

Li Lin Yang

Invariant-mass distribution of top-quark pairs and top-quark mass determination Supported in part by the National Natural Science Foundation of China (11975030, 11635001, 11575004). W.-L. Ju was Supported in part by the China Postdoctoral Science Foundation (2017M610685)

The prompt and non-prompt production cross-sections for ψ (2S) mesons are measured using 2.1 fb −1 of pp collision data at a centre-of-mass energy of 7 TeV recorded by the ATLAS experiment at the LHC. The measurement exploits the ψ (2S) → J/ψ (→ μ + μ − ) π + π − decay mode, and probes ψ (2S) mesons with transverse momenta in the range 10 ≤ p T &lt; 100 GeV and rapidity | y | &lt; 2 . 0. The results are compared to other measurements of ψ (2S) production at the LHC and to various theoretical models for prompt and non-prompt quarkonium production.

France

Georges Aad

Brad Abbott

Jalal Abdallah

Samah Abdel Khalek

Azerbaijan

Ovsat Abdinov

Netherlands

Rosemarie Aben

Babak Abi

Maris Abolins

Canada

Ossama AbouZeid

Israel

Halina Abramowicz

Henso Abreu

CERN

Ricardo Abreu

Sweden

Yiming Abulaiti

Bobby Samir Acharya

Leszek Adamczyk

David Adams

Jahred Adelman

Stefanie Adomeit

Tim Adye

Montenegro

Tatjana Agatonovic-Jovin

Spain

Juan Antonio Aguilar-Saavedra

Marco Agustoni

Steven Ahlen

Faig Ahmadov

Italy

Giulio Aielli

Henrik Akerstedt

Torsten Paul Ake Åkesson

Japan

Ginga Akimoto

Russia

Andrei Akimov

Gian Luigi Alberghi

Justin Albert

Solveig Albrand

Argentina

Maria Josefina Alconada Verzini

Martin Aleksa

Igor Aleksandrov

Romania

Calin Alexa

Gideon Alexander

Gauthier Alexandre

Greece

Theodoros Alexopoulos

Muhammad Alhroob

Gianluca Alimonti

Lion Alio

John Alison

Benedict Allbrooke

Lee John Allison

Phillip Allport

John Almond

Alberto Aloisio

Denmark

Alejandro Alonso

Francisco Alonso

Cristiano Alpigiani

Andrew David Altheimer

Barbara Alvarez Gonzalez

Mariagrazia Alviggi

Katsuya Amako

Brazil

Yara Amaral Coutinho

Christoph Amelung

Dante Amidei

Portugal

Susana Patricia Amor Dos Santos

Antonio Amorim

Simone Amoroso

Nir Amram

Glenn Amundsen

Christos Anastopoulos

Lucian Stefan Ancu

Nansi Andari

Timothy Andeen

Christoph Falk Anders

Gabriel Anders

Kelby Anderson

Attilio Andreazza

George Victor Andrei

Xabier Anduaga

Stylianos Angelidakis

Ivan Angelozzi

Philipp Anger

Aaron Angerami

Francis Anghinolfi

Alexey Anisenkov

Nuno Anjos

Alberto Annovi

Ariadni Antonaki

Mario Antonelli

Alexey Antonov

Slovakia

Jaroslav Antos

Fabio Anulli

Masato Aoki

Ludovica Aperio Bella

Rudi Apolle

Giorgi Arabidze

Ignacio Aracena

Yasuo Arai

Juan Pedro Araque

Ayana Arce

Jean-Francois Arguin

Spyridon Argyropoulos

Türkiye

Metin Arik

Aaron James Armbruster

Olivier Arnaez

Vanessa Arnal

Measurement of the production cross-section of ψ (2S) → J/ψ (→ μ + μ − ) π + π − in pp collisions at s $$ \sqrt{s} $$ = 7 TeV at ATLAS