Michael Kagan
Stanford University
H-index: 194
North America-United States
Description
Michael Kagan, With an exceptional h-index of 194 and a recent h-index of 140 (since 2020), a distinguished researcher at Stanford University, specializes in the field of Particle Physics, Machine Learning.
His recent articles reflect a diverse array of research interests and contributions to the field:
Re-Simulation-based Self-Supervised Learning for Pre-Training Foundation Models
A Faster Way To Yes: Re-Balancing American Asylum Procedures
Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models
Laser Wavefront Engineering and Metrology using Point Source Atom Interferometry with 3-D Imaging Reconstruction
Terrestrial very-long-baseline atom interferometry: Workshop summary
Differentiable matrix elements with MadJax
Branches of a Tree: Taking Derivatives of Programs with Discrete and Branching Randomness in High Energy Physics
Systems, devices and methods for ultra-sensitive detection of molecules or particles
Professor Information
University | Stanford University |
---|---|
Position | Panofsky Fellow SLAC |
Citations(all) | 206940 |
Citations(since 2020) | 95613 |
Cited By | 151841 |
hIndex(all) | 194 |
hIndex(since 2020) | 140 |
i10Index(all) | 763 |
i10Index(since 2020) | 667 |
University Profile Page | Stanford University |
Research & Interests List
Particle Physics
Machine Learning
Top articles of Michael Kagan
Re-Simulation-based Self-Supervised Learning for Pre-Training Foundation Models
Self-Supervised Learning (SSL) is at the core of training modern large machine learning models, providing a scheme for learning powerful representations that can be used in a variety of downstream tasks. However, SSL strategies must be adapted to the type of training data and downstream tasks required. We propose RS3L, a novel simulation-based SSL strategy that employs a method of re-simulation to drive data augmentation for contrastive learning. By intervening in the middle of the simulation process and re-running simulation components downstream of the intervention, we generate multiple realizations of an event, thus producing a set of augmentations covering all physics-driven variations available in the simulator. Using experiments from high-energy physics, we explore how this strategy may enable the development of a foundation model; we show how R3SL pre-training enables powerful performance in downstream tasks such as discrimination of a variety of objects and uncertainty mitigation. In addition to our results, we make the RS3L dataset publicly available for further studies on how to improve SSL strategies.
Authors
Philip Harris,Michael Kagan,Jeffrey Krupa,Benedikt Maier,Nathaniel Woodward
Journal
arXiv preprint arXiv:2403.07066
Published Date
2024/3/11
A Faster Way To Yes: Re-Balancing American Asylum Procedures
The United States asylum system, like many other asylum systems, is under immense pressure to process asylum applications faster. The primary response to this pressure is negative, namely to deny asylum claims quickly because they are manifestly unfounded. In the US, this is done through the credible fear process. This negative orientation leads to a structural imbalance in which denials can be fast and easy for the system, but approvals take time and extensive effort. Using domestic and international comparative examples, this Article proposes re-balancing the asylum system by establishing a process for expedited approvals of clearly eligible asylum claims.
Authors
Michael Kagan
Journal
Georgetown Law Journal, Forthcoming
Published Date
2024/3/9
Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models
We propose \textit{masked particle modeling} (MPM) as a self-supervised method for learning generic, transferable, and reusable representations on unordered sets of inputs for use in high energy physics (HEP) scientific data. This work provides a novel scheme to perform masked modeling based pre-training to learn permutation invariant functions on sets. More generally, this work provides a step towards building large foundation models for HEP that can be generically pre-trained with self-supervised learning and later fine-tuned for a variety of down-stream tasks. In MPM, particles in a set are masked and the training objective is to recover their identity, as defined by a discretized token representation of a pre-trained vector quantized variational autoencoder. We study the efficacy of the method in samples of high energy jets at collider physics experiments, including studies on the impact of discretization, permutation invariance, and ordering. We also study the fine-tuning capability of the model, showing that it can be adapted to tasks such as supervised and weakly supervised jet classification, and that the model can transfer efficiently with small fine-tuning data sets to new classes and new data domains.
Authors
Lukas Heinrich,Michael Kagan,Samuel Klein,Matthew Leigh,Tobias Golling,John Andrew Raine,Margarita Osadchy
Journal
arXiv preprint arXiv:2401.13537
Published Date
2024/1/24
Laser Wavefront Engineering and Metrology using Point Source Atom Interferometry with 3-D Imaging Reconstruction
R02. 00005: Laser Wavefront Engineering and Metrology using Point Source Atom Interferometry with 3-D Imaging Reconstruction
Authors
Yiping Wang,Sean Gasiorowski,Michael Kagan,Tim Kovachy
Journal
Bulletin of the American Physical Society
Published Date
2024/6/6
Terrestrial very-long-baseline atom interferometry: Workshop summary
This document presents a summary of the 2023 Terrestrial Very-Long-Baseline Atom Interferometry Workshop hosted by CERN. The workshop brought together experts from around the world to discuss the exciting developments in large-scale atom interferometer (AI) prototypes and their potential for detecting ultralight dark matter and gravitational waves. The primary objective of the workshop was to lay the groundwork for an international TVLBAI proto-collaboration. This collaboration aims to unite researchers from different institutions to strategize and secure funding for terrestrial large-scale AI projects. The ultimate goal is to create a roadmap detailing the design and technology choices for one or more km-scale detectors, which will be operational in the mid-2030s. The key sections of this report present the physics case and technical challenges, together with a comprehensive overview of the discussions at the workshop together with the main conclusions.
Authors
Sven Abend,Baptiste Allard,Iván Alonso,John Antoniadis,Henrique Araujo,Gianluigi Arduini,Aidan Arnold,Tobias Aßmann,Nadja Augst,Leonardo Badurina,Antun Balaz,Hannah Banks,Michele Barone,Michele Barsanti,Angelo Bassi,Baptiste Battelier,Charles Baynham,Beaufils Quentin,Aleksandar Belic,Ankit Beniwal,Jose Bernabeu,Francesco Bertinelli,Andrea Bertoldi,Ikbal Ahamed Biswas,Diego Blas,Patrick Boegel,Aleksandar Bogojevic,Jonas Böhm,Samuel Böhringer,Kai Bongs,Philippe Bouyer,Christian Brand,Apostolos Brimis,Oliver Buchmueller,Luigi Cacciapuoti,Sergio Calatroni,Benjamin Canuel,Chiara Caprini,Ana Caramete,Laurentiu Caramete,Matteo Carlesso,John Carlton,Mateo Casariego,Vassilis Charmandaris,Yu-Ao Chen,Maria Luisa Chiofalo,Alessia Cimbri,Jonathon Coleman,Florin Lucian Constantin,Carlo Contaldi,Yanou Cui,Elisa Da Ros,Gavin Davies,Esther del Pino Rosendo,Christian Deppner,Andrei Derevianko,Claudia de Rham,Albert De Roeck,Daniel Derr,Fabio Di Pumpo,Goran Djordjevic,Babette Dobrich,Peter Domokos,Peter Dornan,Michael Doser,Giannis Drougakis,Jacob Dunningham,Alisher Duspayev,Sajan Easo,Joshua Eby,Maxim Efremov,Tord Ekelof,Gedminas Elertas,John Ellis,David Evans,Pavel Fadeev,Mattia Fanì,Farida Fassi,Marco Fattori,Pierre Fayet,Daniel Felea,Jie Feng,Alexander Friedrich,Elina Fuchs,Naceur Gaaloul,Dongfeng Gao,Susan Gardner,Barry Garraway,Alexandre Gauguet,Sandra Gerlach,Matthias Gersemann,Valerie Gibson,Enno Giese,Gian Francesco Giudice,Eric Glasbrenner,Mustafa Gündogan,Martin G Haehnelt,Timo Hakulinen,Klemens Hammerer,Ekim Taylan Hanımeli,Tiffany Harte,Leonie Hawkins,Aurelien Hees,Jaret Heise,Victoria Henderson,Sven Herrmann,Thomas Hird,Jason Hogan,Bodil Holst,Michael Holynski,Kamran Hussain,Gregor Janson,Peter Jeglič,Fedor Jelezko,Michael Kagan,Matti Kalliokoski,Mark Kasevich,Alex Kehagias,Eva Kilian,Soumen Koley,Bernd Konrad,Joachim Kopp,Georgy Kornakov,Tim Kovachy,Markus Krutzik,Mukesh Kumar,Pradeep Kumar,Claus Laemmerzahl,Greg Landsberg,Mehdi Langlois,Bryony Lanigan,Samuel Lellouch,Bruno Leone,Christophe Le Poncin Lafitte,Marek Lewicki,Bastian Leykauf,Ali Lezeik,Lucas Lombriser,Luis López,Elias López Asamar,Cristian López Monjaraz,Gaetano Luciano,Mohammed Mahmoud Mohammed,Azadeh Maleknejad,Krutzik Markus,Jacques Marteau,Didier Massonnet,Anupam Mazumdar,Christopher McCabe,Matthias Meister
Journal
arXiv preprint arXiv:2310.08183
Published Date
2023/10/12
Differentiable matrix elements with MadJax
MadJax is a tool for generating and evaluating differentiable matrix elements of high energy scattering processes. As such, it is a step towards a differentiable programming paradigm in high energy physics that facilitates the incorporation of high energy physics domain knowledge, encoded in simulation software, into gradient based learning and optimization pipelines. MadJax comprises two components:(a) a plugin to the general purpose matrix element generator MadGraph that integrates matrix element and phase space sampling code with the JAX differentiable programming framework, and (b) a standalone wrapping code interface for accessing the matrix element code and its gradients, which are computed with automatic differentiation. The MadJax implementation and example applications of simulation based inference and normalizing flow based matrix element modeling, with capabilities enabled uniquely …
Authors
Lukas Heinrich,Michael Kagan
Journal
Journal of Physics: Conference Series
Published Date
2023/2/1
Branches of a Tree: Taking Derivatives of Programs with Discrete and Branching Randomness in High Energy Physics
We propose to apply several gradient estimation techniques to enable the differentiation of programs with discrete randomness in High Energy Physics. Such programs are common in High Energy Physics due to the presence of branching processes and clustering-based analysis. Thus differentiating such programs can open the way for gradient based optimization in the context of detector design optimization, simulator tuning, or data analysis and reconstruction optimization. We discuss several possible gradient estimation strategies, including the recent Stochastic AD method, and compare them in simplified detector design experiments. In doing so we develop, to the best of our knowledge, the first fully differentiable branching program.
Authors
Michael Kagan,Lukas Heinrich
Journal
arXiv preprint arXiv:2308.16680
Published Date
2023/8/31
Systems, devices and methods for ultra-sensitive detection of molecules or particles
UPSFMJHZUCSEHU-JYGUBCOQSA-N n-[(2s, 3r, 4r, 5s, 6r)-2-[(2r, 3s, 4r, 5r, 6s)-5-acetamido-4-hydroxy-2-(hydroxymethyl)-6-(4-methyl-2-oxochromen-7-yl) oxyoxan-3-yl] oxy-4, 5-dihydroxy-6-(hydroxymethyl) oxan-3-yl] acetamide Chemical compound CC (= O) N [C@@ H] 1 [C@@ H](O)[C@ H](O)[C@@ H](CO) O [C@ H] 1O [C@ H] 1 [C@ H](O)[C@@ H](NC (C)= O)[C@ H](OC= 2C= C3OC (= O) C= C (C) C3= CC= 2) O [C@@ H] 1CO UPSFMJHZUCSEHU-JYGUBCOQSA-N 0.000 description 3
Published Date
2023/1/18
Professor FAQs
What is Michael Kagan's h-index at Stanford University?
The h-index of Michael Kagan has been 140 since 2020 and 194 in total.
What are Michael Kagan's top articles?
The articles with the titles of
Re-Simulation-based Self-Supervised Learning for Pre-Training Foundation Models
A Faster Way To Yes: Re-Balancing American Asylum Procedures
Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models
Laser Wavefront Engineering and Metrology using Point Source Atom Interferometry with 3-D Imaging Reconstruction
Terrestrial very-long-baseline atom interferometry: Workshop summary
Differentiable matrix elements with MadJax
Branches of a Tree: Taking Derivatives of Programs with Discrete and Branching Randomness in High Energy Physics
Systems, devices and methods for ultra-sensitive detection of molecules or particles
...
are the top articles of Michael Kagan at Stanford University.
What are Michael Kagan's research interests?
The research interests of Michael Kagan are: Particle Physics, Machine Learning
What is Michael Kagan's total number of citations?
Michael Kagan has 206,940 citations in total.