ABOUT

MODE (for Machine-learning Optimized Design of Experiments) is a collaboration of physicists and computer scientists who target the use of differentiable programming in design optimization of detectors for particle physics applications, extending from fundamental research at accelerators, in space, and in nuclear physics and neutrino facilities, to industrial applications employing the technology of radiation detection.
We aim to develop modular, customizable, and scalable, fully differentiable pipelines for the end-to-end optimization of articulated objective functions that model in full the true goals of experimental particle physics endeavours, to ensure optimal detector performance, analysis potential, and cost-effectiveness.

The main goal of our activities is to develop an architecture that can be adapted to the above use cases but will also be customizable to any other experimental endeavour employing particle detection at its core. We welcome suggestions, as well as interest in joining our effort, by researchers focusing on use cases for which this technology can be of benefit.

The above program is supported as an expression of interest by the JENAA group.

WORKSHOP

The Fourth MODE Workshop on Differentiable Programming for experiment design will take place at IFIC Valencia! Registration and abstract submission are already open in the indico page!. Keynote talks by Danilo Rezende (DeepMind), Andrea Walther (HUBerlin), and Riccardo Zecchina (Bocconi).

NEWS

2024

• May 13: the preprint Optimization Using Pathwise Algorithmic Derivatives of Electromagnetic Shower Simulations/a>, led by several MODE members, is now online!!! Congratulations to all involved!.
• May 7: the paper TomOpt: Differential optimisation for task- and constraint-aware design of particle detectors in the context of muon tomography has been accepted for publication by "Machine Learning: Science and Technology"!!! Congratulations to all involved!.
• Mar 01: the Fourth MODE Workshop on Differentiable Programming for experiment design will take place at IFIC Valencia! Registration and abstract submission are already open in the indico page!. Keynote talks by Danilo Rezende (DeepMind), Andrea Walther (HUBerlin), and Riccardo Zecchina (Bocconi).

2023

• Oct 3: the preprint End-To-End Optimization of the Layout of a Gamma Ray Observatory, led by several MODE members, is now online!!! Congratulations to all involved!.
• Sep 25: our preprint TomOpt: Differential optimisation for task- and constraint-aware design of particle detectors in the context of muon tomography is now online!!! Congratulations to all involved!.
• May 30: several members of MODE (Aehle, Baydin, Gauger, Haack, Heinrich, Sagebaum, Vischia, Watts) will be present at Differentiable and Probabilistic Programming for Fundamental Physics long research workshop in Munich next month!
• May 25: our white paper Toward the end-to-end optimization of particle physics instruments with differentiable programming is now published!!! Congratulations to all involved!.
• May 18: the first confirmed contributions to the Third MODE Workshop on Differentiable Programming for experiment design are published in the agenda!.
• Mar 20: the Third MODE Workshop on Differentiable Programming for experiment design will take place at Princeton University! Registration and abstract submission are already open in the indico page!.

2022

• Dec 8: the contribution by MODE (and collaborators) to the NuPECC Long Range Plan 2024 has been made public in NuPECC's dedicated webpage.
• Nov 25: a collaboration meeting of MODE was held, and seven new members were accepted.
• Sep 12-16: The Second MODE Workshop was held in Kolymbari, Crete, with 50 in-person participants and 36 online participants.
• May 20: registrations and abstract submission are now open for the Second MODE Workshop on Differentiable Programming for Experiment Design.
• Mar 22: A white paper by (among others) MODE authors, "Toward the End-to-End Optimization of Particle Physics Instruments with Differentiable Programming: a White Paper", has been uploaded to the arXiv. See our Publications section or the arXiv.
• Mar 10: An article by (among others) MODE authors, "neos: End-to-End-Optimised Summary Statistics for High Energy Physics", has been uploaded to the arXiv. See our Publications section or the arXiv.
• Mar 6: An article by (among others) MODE authors, "Deep Regression of Muon Energy with a K-Nearest Neighbor Algorithm", has been uploaded to the arXiv. See our Publications section or the arXiv.
• Feb 28: An article by (among others) MODE authors, "Differentiable Matrix Elements with MadJax", has been uploaded to the arXiv. See our Publications section or the arXiv.
• Feb 22: An article by (among others) MODE authors, "Calorimetric Measurement of Multi-TeV Muons via Deep Regression", has been published in EPJ-C! See our Publications section or the journal website

2021

• Oct 21: The MODE Collaboration has voted in favour of several candidatures: we are pleased to welcome in the Collaboration: Dr. Tasos Belias (GSI), Dr. Claudius Krause (Rutgers), Prof. Christian Glaser (Uppsala), Mr. Max Lamparth (TUM), Dr. Lukas Heinrich (CERN), Dr. Patrick Stowell (Durham), and Prof. Haitham Zaraket (Lebanese University)!
• Sep 9: The first MODE workshop on automatic differentiation has been a success! More than 100 participants, about one fourth of which attended in person, gathered to discuss automatic differentiation in several research fields, and a whitepaper is coming!
• Jul 22: External registrations are now open for the first MODE workshop on automatic differentiation! Follow the link to the registration page contained in our workshop page!!!
• May 14: an organizational meeting between INFN researchers interested in the MODE research program has taken place. A new group "MODE_INFN" will be created within INFN Group 5, with the aim of enabling a collaboration of INFN researchers in Padova, Firenze, and Naples, as well as international researchers, for software development and hardware studies within the areas of muon tomography and calorimetry, plus maybe others.
• Apr. 25: the ParticleBites blog has put online a post on MODE.
• Apr. 21: the APPEC website and newsletter feature an interview to the Scientific Coordinator of MODE, Dr. Tommaso Dorigo
• Mar. 30: the first article by the MODE Collaboration is published in Nuclear Physics News International, and can be consulted in the journal website
• Mar. 10: the preprint of our short article to be published by Nuclear Physics News International is now available on INSPIRE
• Mar. 1: we have made public on our website the page for the first MODE Workshop on Differentiable Programming
• Feb. 5: MODE has made public the preprint of the short article that it submitted in December to Nuclear Physics News International. You can read the preprint here
• Feb. 5: we have made public in this page the contacts of the members of the Steering Board and of the representatives of the institutions

2020

• Dec. 20: MODE has submitted a short article to Nuclear Physics News International, where we summarize our research plan. The article will soon be available online.
• Dec. 14: the Scientific Coordinator and the composition of the Steering Board of the MODE Collaboration until Dec. 31 2022 have been approved by the collaboration.
• Nov. 9: The Statute of the MODE Collaboration was approved.
• Oct. 16: At 4PM took place a kick-off meeting of the MODE collaboration.
• Sep. 8: MODE decided to organize a workshop on "Differentiable Programming for Design Optimization", in February or June (depending on the evolution of the COVID-19 situation) in Louvain-la-Neuve (or Padova).
• Sep. 6: An open kick-off meeting of the JENAA group, where we will discuss ways to obtain support of the community to start our activities, is scheduled for September 15, 2-4PM CET. Feel free to participate! The indico page of the meeting is available here.

COLLABORATION

At INFN and Università of Padova Dr. Tommaso Dorigo, Dr. Pablo De Castro Manzano, Dr. Federica Fanzago, Dr. Lukas Layer, Dr. Giles Strong, Dr. Mia Tosi, and Dr. Hevjin Yarar
At Université catholique de Louvain Dr. Andrea Giammanco, Prof. Christophe Delaere, and Mr. Maxime Lagrange
At Universidad de Oviedo and ICTEA Dr. Pietro Vischia
At Université Clermont Auvergne, Prof. Julien Donini, and Mr. Federico Nardi (joint with Universitá di Padova)
At the Higher School of Economics of Moscow, Prof. Andrey Ustyuzhanin, Dr. Alexey Boldyrev, Dr. Denis Derkach, and Dr. Fedor Ratnikov
At the Instituto de Física de Cantabria, Dr. Pablo Martínez Ruíz del Árbol
At CERN, Dr. Sofia Vallecorsa
At Karlsruher Institut für Technologie, Dr. Jan Kieseler
At University of Oxford Dr. Atilim Gunes Baydin
At New York University Prof. Kyle Cranmer
At Université de Liège Prof. Gilles Louppe
At GSI/FAIR Dr. Anastasios Belias
At HEPHY Vienna (OeAW) Dr. Claudius Krause
At Uppsala Universitet Prof. Christian Glaser
At TU-München, Prof. Lukas Heinrich and Mr. Max Lamparth
At Durham University Dr. Patrick Stowell
At Lebanese University Prof. Haitham Zaraket
At University of Kaiserslautern-Landau Mr. Max Aehle, Prof. Nicolas Gauger, Dr. Lisa Kusch
At University of Applied Sciences Worms Prof. Ralf Keidel
At Princeton University Prof. Peter Elmer
At University of Washington Prof. Gordon Watts
At SLAC Dr. Ryan Roussel
At Lulea University of Technology Prof. Fredrik Sandin and Prof. Marcus Liwicki
At IGFAE and Universidad de Santiago de Compostela Prof. Xabier Cid Vidal

The Scientific Coordinator of the MODE Collaboration is Dr. Tommaso Dorigo, INFN-Sezione di Padova
The Steering Board of the MODE Collaboration includes:

• Prof. Julien Donini, UCA
• Dr. Tommaso Dorigo, INFN-PD
• Dr. Andrea Giammanco, UCLouvain
• Dr. Fedor Ratnikov, HSE
• Dr. Pietro Vischia, UniOvi

EVENTS

Below is a list of events organized by the MODE Collaboration:

• Fourth MODE Workshop on Differentiable Programming for Experiment Design (Sept 23-25 2024, Valencia, Spain): the workshop page is here REGISTRATIONS ARE OPEN!.
• TomNeutron hackathon (Mar 14-15, 2024), Santiago de Compostela, Spain.
• TomOpt hackathon (Feb 21-23, 2024), Oviedo, Spain.
• Third MODE Workshop on Differentiable Programming for Experiment Design (July 24-26 2023, Princeton, USA): the workshop page is here.
• TomOpt workshop (Nov 28 - Dec 2, 2022), Louvain-la-Neuve, Belgium.
• Second MODE Workshop on Differentiable Programming for Experiment Design (Sep 12-16 2022, Kolymbari, Crete): the workshop page is here.
• First MODE Workshop on Differentiable Programming (Sep 3-6 2021, Louvain-la-Neuve, Belgium): the workshop page is now available on our website

PUBLICATIONS

Below is a list of publications by the MODE Collaboration:

• M. Aehle et al. (several MODE authors), "Plans for End-to-End Optimization of Experiments", plan submitted to the NuPECC Long Range Plan 2024.
• T. Dorigo, A. Giammanco, P. Vischia (editors) et al., "Toward the End-to-End Optimization of Particle Physics Instruments with Differentiable Programming: a White Paper", published in Reviews in Physics 10 (2023) 100085. Available in the journal website or as a preprint, 2203.13818.
• The MODE Collaboration, "Toward Machine Learning Optimization of Experimental Design", published in Nuclear Physics News International in March 2021 (submitted December 2020). Available in the journal website or as a preprint here and on INSPIRE

Below is a concise list of relevant publications to the research interests of the MODE Collaboration. MODE members among the authors are indicated in boldface:

• Max Aehle, Mihály Novák, Vassil Vassilev, Nicolas R. Gauger, Lukas Heinrich, Michael Kagan, David Lange, "Optimization Using Pathwise Algorithmic Derivatives of Electromagnetic Shower Simulations", arXiv:2405.07944
• Tommaso Dorigo, Max Aehle, Julien Donini, Michele Doro, Nicolas R. Gauger, Rafael Izbicki, Ann Lee, Luca Masserano, Federico Nardi, Sidharth S S, Alexander Shen, "End-To-End Optimization of the Layout of a Gamma Ray Observatory", arXiv:2310.01857
• Giles C. Strong, Maxime Lagrange, Aitor Orio, Anna Bordignon, Florian Bury, Tommaso Dorigo, Andrea Giammanco, Mariam Heikal, Jan Kieseler, Max Lamparth, Pablo Martínez Ruíz del Árbol, Federico Nardi, Pietro Vischia, Haitham Zaraket, "TomOpt: Differential optimisation for task- and constraint-aware design of particle detectors in the context of muon tomography", arXiv:2309.14027 (accepted by "Machine Learning: Science and Technology")
• Michael Kagan, L. Heinrich, "Branches of a Tree: Taking Derivatives of Programs with Discrete and Branching Randomness in High Energy Physics", arXiv:2308.16680
• N. Simpson, L. Heinrich, "neos: End-to-End-Optimised Summary Statistics for High Energy Physics", arXiv:2203.05570
• T. Dorigo, S. Guglielmini, J. Kieseler, L. Layer, G.C. Strong, "Deep Regression of Muon Energy with a K-Nearest Neighbor Algorithm", arXiv:2203.02841.
• L. Heinrich, M. Kagan, "Differentiable Matrix Elements with MadJax", arXiv:2203.00057
• J. Kieseler, G.C. Strong, F. Chiandotto, T. Dorigo, L. Layer, "Calorimetric Measurement of Multi-TeV Muons via Deep Regression" Eur. Phys. J. C (2022) 82: 79, doi:10.1140/epjc/s10052-022-09993-5
• C. Neubüser, Jan Kieseler, Paul Lujan, "Optimising longitudinal and lateral calorimeter granularity for software compensation in hadronic showers using deep neural networks" Eur. Phys. J. C (2022) 82: 92 (2022) doi:10.1140/epjc/s10052-022-10031-7
• A. Gunes Baydin, B.A. Pearlmutter, A.A. Radul, and J.M. Siskind, "Automatic Differentiation in Machine Learning: a Survey", Journal of Machine Learning Research (JMLR) 18 (153) (2018) 1, http://jmlr.org/papers/v18/17-468.html
• T. Dorigo, "Geometry Optimization of a Muon-Electron Scattering Detector," Physics Open 4 (2020) 100022, arXiv:200200973[physics.ins-det], doi:10.1016/j.physo.2020.100022.
• T. Dorigo, J. Kieseler, L. Layer and G. Strong, "Muon Energy Measurement from Radiative Losses in a Calorimeter for a Collider Detector", arXiv:2008.10958 [physics.ins-det] (2020).
• S. Shirobokov, A. Ustyuzhanin, A. Gunes Badyin et al., "Differentiating the Black-Box: Optimization with Local Generative Surrogates", arXiv:2002.04632v1 [cs.LG] (2020).
• K. Cranmer, J. Brehmer, and G. Louppe, "The frontier of simulation-based inference", arXiv:1911.01429[stat.ML] (2019), Proceedings of the National Academy of Sciences.
• F. Ratnikov, "Using machine learning to speed up and improve calorimeter R&D", JINST 15 (2020) C05032, doi: 10.1088/1748-0221/15/05/C05032.
• F. Ratnikov, D. Derkach, A. Boldyrev, A. Shevelev, P. Fakanov, L. Matyushin, "Using machine learning to speed up new and upgrade detector studies: a calorimeter case", to appear in proceedings of CHEP 2019, https://arxiv.org/abs/2003.05118
• A. Boldyrev, D. Derkach, F. Ratnikov, A. Shevelev, "ML-assisted versatile approach to Calorimeter R&D", arXiv:2005.07700
• P. Giubilato et al., "iMPACT: innovative pCT scanner", IEEE Nucl. Science Symposium and Medical Imaging Conference (NSS/MIC) IEEE (2015), https://ieeexplore.ieee.org/abstract/document/7581240.
• S. Wuyckens, A. Giammanco, P. Demin, and E. Cortina Gil, "A Portable muon telescope based on small and gas-tight Resistive Plate Chambers"<, Phil. Trans. Royal Soc. A377 (2019) 2137, arXiv:1806.06602v2[physics.ins-det] (2018), doi:10.1098/rsta.2018.0139.
• J. Kieseler, "Object condensation: one-stage grid-free multi-object reconstruction in physics detectors, graph and image data", arXiv:2002.03605[physics.data-an] (2020).
• P. de Castro Manzano and T. Dorigo, "INFERNO: Inference-Aware Neural Optimization", Comp. Phys. Commun. 244 (2019) 170; Arxiv:1806.04743v2 [stat.ml] (2018), doi:10.1016/j.cpc.2019.06.007 .
• J. Brehmer, K. Cranmer et al., "MadMiner: Machine learning-based inference for particle physics", Comput. Softw. Big Sci. 4 (2020) 1, 3, doi:10.1007/s41781-020-0035-2.
• G. Louppe, J. Hermans, and K. Cranmer, "Adversarial Variational Optimization of Non-Differentiable Simulators", PMLR 89:1438-1447, 2019, arXiv:1707.07113[stat.ML].
• K. Cranmer, J. Pavez, and G. Louppe, "Approximating Likelihood Ratios with Calibrated Discriminative Classifiers", arXiv:1506.02169[stat.ML] (2015).

CONTACTS

Scientific Coordinator	Dr. Tommaso Dorigo Email: dorigo (at) pd (dot) infn (dot) it Office tel.: +39 0499677230 Mobile tel.: +39 3666995594 Snail mail: Dipartimento di Fisica e Astronomia "G.Galilei", via Francesco Marzolo 8, 35131 Padova, Italy
Other Steering Board members	Prof. Julien Donini, julien (dot) donini (at) cern (dot) ch Dr. Andrea Giammanco, andrea (dot) giammanco (at) uclouvain (dot) be Dr. Fedor Ratnikov, fratnikov (at) hse (dot) ru Dr. Pietro Vischia, pietro (dot) vischia (at) cern (dot) ch
Institutional contacts (when not in SB already and for multimember institutions):	CERN: Dr. Jan Kieseler, jan (dot) kieseler (at) cern (dot) ch Université de Liège: Prof. Gilles Louppe, g (dot) louppe (at) uliege (dot) be TU-München: Prof. Lukas Heinrich, lukas (dot) heinrich (at) cern (dot) ch Technische Universitàët Kaiserslautern: Prof. Nicolas Gauger, nicolas (dot) gauger (at) scicomp (dot) uni-kl (dot) de (the email address of some institutional contacts is not disclosed)