ML4Jets2024

from Monday 4 November 2024 (08:30) to Friday 8 November 2024 (17:00)
LPNHE, Paris, France

        : Sessions
    /     : Talks
        : Breaks
4 Nov 2024
5 Nov 2024
6 Nov 2024
7 Nov 2024
8 Nov 2024
AM
08:30
Breaks (until 09:00)
09:00
Welcome & Opening Talks (until 09:40)
09:00 Welcome  
09:40
Plenary talks (until 10:20)
09:40 Symbolic machine learning in physics - Francois Charton (Meta AI Paris)  
10:20 --- Coffee break ---
10:50
Plenary talks (until 12:10)
10:50 Experimental highlights: Edge AI for particle physics - Jennifer Ngadiuba (FNAL)  
11:30 Experimental highlights: The State of ML in LHC Science - Daniel Thomas Murnane (Niels Bohr Institute, University of Copenhagen)  
09:00
Reconstruction - Fabrice Balli (Université Paris-Saclay (FR)) (until 10:20) (Amphi Charpak)
09:00 Deep learning on jet modification in the presence of the QGP background - RAN LI  
09:20 Transformer networks for constituent-based b-jet calibration with the ATLAS detector - Brendon Bullard (SLAC National Accelerator Laboratory (US))  
09:40 Jet Finding as a Real-Time Object Detection Task - Leon Bozianu (Universite de Geneve (CH))  
10:00 Transformer for Energy Calibration in the ATLAS Electromagnetic Calorimeter - Ryan Roberts (University of California Berkeley (US))  
09:00
Theorie - Tilman Plehn (until 10:20) (Salle Séminaires)
09:00 Learning the Simplicity of Scattering Amplitudes - Aurelien Dersy (Harvard University)  
09:20 Taming perturbation theory in QCD with Normalizing Flows - Rikab Gambhir (MIT)  
09:40 Efficient SMEFT fits with neural importance sampling - Nikita Schmal  
10:00 Machine learning the likelihoods - Rafal Maselek  
10:20 --- Coffee break ---
10:50
Event generation - Ramon Winterhalder (Università degli Studi di Milano) (until 12:10) (Salle séminaires)
10:50 Exploring phase space with Flow Matching - Timo Janssen  
11:10 Differentiable MadNIS-Lite - Theo Heimel (Heidelberg University)  
11:30 Event Generation with Lorentz-Equivariant Geometric Algebra Transformers - Jonas Simon Spinner  
11:50 Classifying importance regions in Monte Carlo simulations with machine learning - Raymundo Ramos (Korea Institute for Advanced Study)  
10:50
Reconstruction - Brendon Bullard (SLAC National Accelerator Laboratory (US)) (until 12:10) (Amphi Charpak)
10:50 Synergizing Physics: Deep Learning Techniques for Time-of-Flight Reconstruction and Jet Tagging in High Energy Physics - Konrad Helms (Georg-August-Universität Göttingen)  
11:10 Particle flow and flavor tagging with DNN for Higgs factories - Taikan Suehara (ICEPP, The University of Tokyo (JP))  
09:00
Detector Simulation - Sascha Diefenbacher (Lawrence Berkeley National Lab. (US)) (until 10:20) (Salle séminaire)
09:00 Fast Perfekt: Regression-based refinement of fast simulation - Lars Stietz (Hamburg University of Technology (DE))  
09:20 Refining CMS Fast Simulation with ML-based regression - Samuel Louis Bein (Universite Catholique de Louvain (UCL) (BE))  
09:40 Faster Than Fast: Pushing the Limits of Simulation with Generative Models in HEP - Sitian Qian (Peking University (CN))  
10:00 Parnassus: An Automated Approach to Accurate, Precise, and Fast Detector Simulation and Reconstruction - Dmitrii Kobylianskii (Weizmann Institute of Science (IL))  
09:00
Tagging -Prof. Mihoko Nojiri (Theory Center, IPNS, KEK) (until 10:20) (amphi Charpak)
09:00 Integrating Energy Flow Networks with Jet Substructure Observables for Enhanced Jet Quenching Studies - João A. Gonçalves (LIP - IST)  
09:20 Jet Charge Classifiers - Rabia Husain  
09:40 WOTAN: Weakly-supervised Optimal Transport Attention-based Noise Mitigation - Nathan Suri Jr. (Yale University (US))  
10:00 Supervised CWoLa: train supervised classifiers without background simulation - Stephen Mulligan (Universite de Genève)  
10:20 --- Coffee break ---
10:50
Tagging - Eilam Gross (Weizmann Institute of Science (IL)) (until 12:10) (Amphi Charpak)
10:50 Hadronic Top Quark Polarimetry with ParticleNet - Zhongtian Dong (University of Kansas)  
11:10 Pretrained event classification model for collider experiments - Joshua Anthony Ho (Lawrence Berkeley National Lab. (US))  
11:30 Enhancing generalization in high energy physics using white-box adversarial attacks - Franck Rothen (Universite de Geneve (CH))  
10:50
Uncertainties & Interpretability - Martin Klassen (Tufts University (US)) (until 12:10) (Salle Séminaires)
10:50 Frequentist Uncertainties on Density Ratios with Ensembles - Sean Benevedes (Massachusetts Institute of Technology)  
11:10 Machine Learning the Top Mass - Katherine Fraser (Harvard University)  
11:30 Calibrating ATLAS calorimeter signals using an uncertainty-aware precision network - Mr Lorenz Vogel (Heidelberg University)  
11:50 Uncertainty Quantification and Anomaly Detection with Evidential Deep Learning - Mark Neubauer (Univ. Illinois at Urbana Champaign (US))  
09:00
Anomaly detection - Barry Dillon (Ulster University) (until 10:20)
09:00 Semi-Supervised Permutation Invariant Particle-Level Anomaly Detection - Gabriel Pinheiro Matos (Columbia University (US))  
09:20 The versatility of flow-based fast calorimeter surrogate models - Ian Pang  
09:40 (R) Lorentz Group Equivariant Autoencoders - Zichun Hao (California Institute of Technology)  
10:00 New Physics Searches with Graph-Based Anomaly Detection in High-Energy Collisions - Ines Isabel Gouveia Cipriano Piedade Moreira (Laboratory of Instrumentation and Experimental Particle Physics (PT))  
09:00
Uncertainties & Interpretability - Manuel Szewc (until 10:20)
09:00 The good, the bad, and the Bayesian? - Nina Elmer  
09:20 Calibrating Bayesian Generative Machine Learning for Bayesiamplification - Sebastian Guido Bieringer (Hamburg University)  
09:40 KAN we improve on HEP classification tasks? Kolmogorov-Arnold Networks applied to an LHC physics example - Florian Alexander Mausolf (Rheinisch Westfaelische Tech. Hoch. (DE))  
10:00 ParticleNet: Calibration of the jet energy scale using pT regression with partial Run3 data collected by the CMS experiment - Matteo Malucchi (ETH Zurich (CH))  
10:20 --- Coffee break ---
10:50
Anomaly detection - Jennifer Ngadiuba (FNAL) (until 12:10) (Amphi Charpak)
10:50 Model-agnostic search for dijet resonances with anomalous jet substructure in proton-proton collisions at $\sqrt{s}$ = 13 TeV with the CMS detector - Aritra Bal (KIT - Karlsruhe Institute of Technology (DE))  
11:10 Efficient Resonant Anomaly Detection - Ranit Das (Rutgers University)  
11:30 TRANSIT your events into a new mass: Fast background interpolation for semi-supervised anomaly detection searches - Mr Ivan Oleksiyuk (UNIGE)  
09:00
Plenary talks - Mark Dayvon Goodsell (Centre National de la Recherche Scientifique (FR)) (until 09:40) (Amphi Charpak)
09:00 Implicit and Explicit Simulation-Based Inference for Cosmology - François LANUSSE  
09:40
Mixed contributions (until 10:20) (Amphi Charpak)
09:40 ML assisted Event Reconstruction in the CMS Phase-2 High Granularity Calorimeter Endcap - Theo Cuisset (LLR / École Polytechnique (FR))  
10:00 Resonant Searches as Cut and Count Experiments - Marie Hein (RWTH Aachen University)  
10:20 --- Coffee break ---
10:50
Plenary talks - Mark Dayvon Goodsell (Centre National de la Recherche Scientifique (FR)) (until 11:30) (Amphi Charpak)
10:50 Theory Overview - Sven-Ludwig Krippendorf (LMU MUNICH)  
11:30
Closing remarks (until 12:00) (Amphi Charpak)
11:30 Closing remarks - Anja Butter (Centre National de la Recherche Scientifique (FR))  
PM
12:10 --- Lunch ---
13:50
Plenary talks - Gregor Kasieczka (Hamburg University (DE)) (until 14:30)
13:50 MadNIS - A journey towards the first ML event generator - Ramon Winterhalder (Università degli Studi di Milano)  
14:30
Mixed contributions - Gregor Kasieczka (Hamburg University (DE)) (until 15:30)
14:30 The Phase Space Distance Between Collider Events - Tianji Cai (SLAC National Accelerator Laboratory)  
14:50 Meta-Learning Quantum Jet Properties with Quantum Generative Models - Yacine Haddad (Northeastern University (US))  
15:10 Introducing Aspen Open Jets: a real-world ML-ready dataset for jet physics - Ian Pang  
15:30 --- Coffee break ---
16:00
Plenary talks - Anja Butter (Centre National de la Recherche Scientifique (FR)) (until 16:40)
16:00 Time Series Anomaly Detection: Overview and New Trends - Paul Boniol (INRIA, ENS)  
16:40
Mixed contributions - Anja Butter (Centre National de la Recherche Scientifique (FR)) (until 17:20)
16:40 The Landscape of Unfolding with Machine Learning - Nathan Huetsch (Heidelberg University, ITP Heidelberg)  
17:00 Fair Universe HiggsML Uncertainty Challenge - Ragansu Chakkappai (IJCLab-Orsay)  
12:10 --- Lunch - Reception at Zamansky tower ---
13:50
Event generation - Anna Hallin (University of Hamburg) (until 15:30)
13:50 Data-driven hadronization models - Manuel Szewc  
14:10 Fast simulation of backgrounds at LHCb - a generalised tool - Alex Marshall (University of Bristol (GB))  
14:30 Generating particle-clouds with discrete features using Markov jump processes - Dr Darius Faroughy (Rutgers University)  
14:50 (R) Application of generative models for full-detector, whole-event simulated event generation and jet background subtraction - Yeonju Go (Brookhaven National Laboratory (US))  
13:50
Tagging - David Shih (until 15:30) (Amphi Charpak)
13:50 The Fundamental Limit of Jet Tagging - Nishank Nilesh Gite (Lawrence Berkeley National Lab. (US))  
14:10 Streamlined jet tagging network assisted by jet prong structure - Prof. Mihoko Nojiri (Theory Center, IPNS, KEK)  
14:30 Learning Symmetry-Independent Jet Representation via Jet-Based Joint Embedding Predictive Architecture - Haoyang Li (Univ. of California San Diego (US))  
14:50 Jet tagging with Lorentz-Equivariant Geometric Algebra Transformers - Víctor Bresó Pla (University of Heidelberg)  
15:30 --- Coffee break ---
16:00
Detector Simulation -Dr Claudius Krause (HEPHY Vienna (ÖAW)) (until 17:40) (Salle Séminaires)
16:00 Generative transformers for learning point-cloud simulations - Henning Rose  
16:20 CaloDREAM -- Detector Response Emulation via Attentive flow Matching - Luigi Favaro  
16:40 Higher Resolution and Angular Conditioning for Normalizing-Flow-based Generation of Calorimeter Showers - Thorsten Lars Henrik Buss (Universität Hamburg (DE))  
17:00 Calo4pQVAE: A calorimeter surrogate for high energy particle-calorimeter interactions using Dwave’s Zephyr topology - J. Quetzalcoatl Toledo-Marin (TRIUMF)  
17:20 BitHEP – Are 1-Bit Networks all we need? - Daohan Wang (HEPHY ÖAW)  
16:00
Tagging - Pierre Antoine Delsart (CNRS / LPSC Grenoble (FR)) (until 17:40) (Amphi Charpak)
16:20 Heavy-Flavour Jet Tagging at LHCb Using Graph Neural Networks - Nathan Grieser (University of Cincinnati (US))  
16:40 Transforming Flavour Tagging on ATLAS - Greta Brianti (CERN)  
17:00 UParT: A unified approach for jet-based object identification in CMS in Run 3 - Uttiya Sarkar (Rheinisch Westfaelische Tech. Hoch. (DE))  
17:20 Performance of the CNN-based tau identification algorithm with Domain Adaptation using Adversarial Machine Learning for Run 2 - Olha Lavoryk (KIT - Karlsruhe Institute of Technology (DE))  
12:10 --- Lunch ---
13:50
Astro & Cosmo - Katia Matcheva (until 15:30) (Salle Séminaires)
13:50 Fusing physics principles and machine learning: inferring dark matter densities of galaxies using stellar catalogs with incomplete kinematic information - Sung Hak Lim (Rutgers University)  
14:10 SKATR -- A Self-Supervised Summary Transformer for the Square Kilometre Array - Ayodele Ore  
14:30 Sweeping the Dust Away: An Unbiased Map of the Milky Way's Dark Matter and Gravitational Potential with Unsupervised Machine Learning - Eric Putney (Rutgers, The State University of New Jersey)  
14:50 SkyCURTAINs: Model agnostic search for Stellar Streams with Gaia data - Stephen Brian Mulligan (Universite de Geneve (CH))  
15:10 Generation of Air Shower Images for Imaging Air Cherenkov Telescopes using Diffusion Models - Christian Elflein (Erlangen Centre for Astroparticle Physics)  
13:50
Reconstruction - Katherine Fraser (Harvard University) (until 15:30) (Amphi Charpak)
13:50 Efficient Particle Tracking and Pileup Mitigation with State space model - Cheng Jiang (The University of Edinburgh (GB))  
14:10 Graph Neural Network-Based Track Finding in the LHCb Vertex Detector - Fotis Giasemis (Centre National de la Recherche Scientifique (FR))  
14:30 Accelerating Graph-based Tracking Tasks with Symbolic Regression - Nathalie Soybelman (Weizmann Institute of Science (IL))  
14:50 Denoising Graph Super-Resolution for Improved Collider Event Reconstruction - Nilotpal Kakati (Weizmann Institute of Science (IL))  
15:10 (R) Generative Neural Networks for Reconstructing Parton-Level Jet Showers after Hadronization - Umar Sohail Qureshi (Vanderbilt University)  
15:30 --- Coffee break ---
16:00
Plenary talks (until 16:40) (Amphi Charpak)
16:00 Realtime reconstruction: Machine learning in reconstruction at LHC - Simon Akar (University of Cincinnati (US))  
16:40
Mixed contributions - Corentin Allaire (IJCLab, Université Paris-Saclay, CNRS/IN2P3) (until 17:30) (Amphi Charpak)
16:40 A Continuous Calibration of the ATLAS Flavor-Tagging Classifiers via Optimal Transportation Maps - Chris Pollard (University of Warwick (GB))  
17:00 The Fast Calorimeter Challenge 2022: Final Evaluation & Lessons Learned - Dr Claudius Krause (HEPHY Vienna (ÖAW))  
19:00 --- 7PM Workshop Dinner at Bouillon Racine ---
12:10 --- Lunch ---
13:50
Detector Simulation - Timo Janssen (until 15:30) (Salle Séminaires)
13:50 A Library for ML-based Fast Calorimeter Shower Simulation at Future Collider Experiments and Beyond - Peter McKeown (CERN)  
14:10 Towards Detector Agnostic Fast Calorimetry Simulation - Piyush Raikwar (CERN)  
14:30 CaloClouds III: Ultra-Fast Geometry-Independent Highly-Granular Calorimeter Simulation - Anatolii Korol  
14:50 Point-Clouds based Diffusion Model on Hadronic Showers - Martina Mozzanica (Hamburg University (DE))  
13:50
Unfolding & Inference - Vinicius Massami Mikuni (Lawrence Berkeley National Lab. (US)) (until 15:30) (Amphi Charpak)
13:50 OmniFoldHI: Advanced ML Unfolding for Heavy-Ion Data - Alexandre Falcão (University of Bergen)  
14:10 Bridging the Generative Unfolding Gap - Sascha Diefenbacher (Lawrence Berkeley National Lab. (US))  
14:30 Measurement of Jet Track Functions with OmniFold-based Binning Corrections in ATLAS Run 2 Data - Jingjing Pan (Yale University (US))  
14:50 Full Event Particle-Level Unfolding with Variable-Length Latent Variational Diffusion - Kevin Thomas Greif (University of California Irvine (US))  
15:10 How to Unfold Top Decays - Sofia Palacios Schweitzer (ITP, University Heidelberg)  
15:30 --- Coffee break ---
16:00
Foundation models - Tianji Cai (SLAC National Accelerator Laboratory) (until 17:40) (Salle Séminaires)
16:00 Learning powerful jet representations via self-supervision - Shudong Wang (Chinese Academy of Sciences (CN))  
16:20 OmniLearn: A Method to Simultaneously Facilitate All Jet Physics Tasks - Vinicius Massami Mikuni (Lawrence Berkeley National Lab. (US))  
16:40 OmniJet-alpha and beyond: foundation model updates - Anna Maria Cecilia Hallin (University of Hamburg)  
17:00 A Novel Approach to Training Foundation Models for Jet-Related Tasks Without Vector Quantization - Masahiro Morinaga (University of Tokyo (JP))  
17:20 Large-Scale Pretraining and Finetuning for Efficient Jet Classification in Particle Physics - Zihan Zhao (Univ. of California San Diego (US))  
16:00
Unfolding & Inference - Chris Pollard (University of Warwick (GB)) (until 17:40) (Amphi Charpak)
16:00 Advanced techniques for SBI in collider physics - Giovanni De Crescenzo (University of Heidelberg)  
16:20 Towards Universal Unfolding using Denoising Diffusion - Martin Klassen (Tufts University (US))  
16:40 Parameter Estimation with Neural Simulation-Based Inference in ATLAS - Mr Arnaud Jean Maury (Université Paris-Saclay (FR))  
17:00 Constraining the Higgs Potential with Neural Simulation-based Inference for Di-Higgs Production - Radha Mastandrea (University of California, Berkeley)  
17:20 Gamma-ray spectrometry of fission fragments : ML analysis of multi-dimensional spectra - Mattéo Ballu