M. Kagan
YOU?
Author Swipe
View article: Log Gaussian Cox Process Background Modeling in High Energy Physics
Log Gaussian Cox Process Background Modeling in High Energy Physics Open
Background modeling is one of the most critical components in high energy physics data analyses, and for smooth backgrounds it is often performed by fitting using an analytic functional form. In this paper a novel method based on Log Gauss…
View article: Simulation-Prior Independent Neural Unfolding Procedure
Simulation-Prior Independent Neural Unfolding Procedure Open
Machine learning allows unfolding high-dimensional spaces without binning at the LHC. The new SPINUP method extracts the unfolded distribution based on a neural network encoding the forward mapping, making it independent of the prior from …
View article: Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future colliders
Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future colliders Open
We demonstrate transfer learning capabilities in a machine-learned algorithm trained for particle-flow reconstruction in high energy particle colliders. This paper presents a cross-detector fine-tuning study, where we initially pretrain th…
View article: Flow annealed importance sampling bootstrap meets differentiable particle physics
Flow annealed importance sampling bootstrap meets differentiable particle physics Open
High-energy physics requires the generation of large numbers of simulated data samples from complex but analytically tractable distributions called matrix elements. Surrogate models, such as normalizing flows, are gaining popularity for th…
View article: Is tokenization needed for masked particle modeling?
Is tokenization needed for masked particle modeling? Open
In this work, we significantly enhance masked particle modeling (MPM), a self-supervised learning scheme for constructing highly expressive representations of unordered sets relevant to developing foundation models for high-energy physics.…
View article: Resimulation-based self-supervised learning for pretraining physics foundation models
Resimulation-based self-supervised learning for pretraining physics foundation models Open
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 ada…
View article: Brief for Amicus Curiae Professors Michael Kagan and Christopher J. Walker in Support of Rehearing En Banc
Brief for Amicus Curiae Professors Michael Kagan and Christopher J. Walker in Support of Rehearing En Banc Open
View article: Optimization using pathwise algorithmic derivatives of electromagnetic shower simulations
Optimization using pathwise algorithmic derivatives of electromagnetic shower simulations Open
Among the well-known methods to approximate derivatives of expectancies computed by Monte-Carlo simulations, averages of pathwise derivatives are often the easiest one to apply. Computing them via algorithmic differentiation typically does…
View article: Flow Annealed Importance Sampling Bootstrap meets Differentiable Particle Physics
Flow Annealed Importance Sampling Bootstrap meets Differentiable Particle Physics Open
High-energy physics requires the generation of large numbers of simulated data samples from complex but analytically tractable distributions called matrix elements. Surrogate models, such as normalizing flows, are gaining popularity for th…
View article: Umami: A Python toolkit for jet flavour tagging
Umami: A Python toolkit for jet flavour tagging Open
View article: Differentiable vertex fitting for jet flavor tagging
Differentiable vertex fitting for jet flavor tagging Open
This work explores the use of differentiable programming to integrate domain knowledge, in the form of domain specific software, into neural networks to develop scientific machine learning systems. We propose a differentiable vertex fittin…
View article: Dataset for flavour tagging R&D
Dataset for flavour tagging R&D Open
This is a dataset for flavour tagging R&D. It consists of b-jets, c-jets and light-jets in equal number and equal distributions of transverse momentum, pseudo-rapidity and track multiplicity. The jets are sampled from ttbar events produced…
View article: Masked particle modeling on sets: towards self-supervised high energy physics foundation models
Masked particle modeling on sets: towards self-supervised high energy physics foundation models Open
We propose 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 n…
View article: Optimization Using Pathwise Algorithmic Derivatives of Electromagnetic Shower Simulations
Optimization Using Pathwise Algorithmic Derivatives of Electromagnetic Shower Simulations Open
Among the well-known methods to approximate derivatives of expectancies computed by Monte-Carlo simulations, averages of pathwise derivatives are often the easiest one to apply. Computing them via algorithmic differentiation typically does…
View article: Re-Simulation-based Self-Supervised Learning for Pre-Training Foundation Models
Re-Simulation-based Self-Supervised Learning for Pre-Training Foundation Models Open
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 ada…
View article: Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models
Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models Open
We propose 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 n…
View article: Differentiable Vertex Fitting for Jet Flavour Tagging
Differentiable Vertex Fitting for Jet Flavour Tagging Open
We propose a differentiable vertex fitting algorithm that can be used for secondary vertex fitting, and that can be seamlessly integrated into neural networks for jet flavour tagging. Vertex fitting is formulated as an optimization problem…
View article: Branches of a Tree: Taking Derivatives of Programs with Discrete and Branching Randomness in High Energy Physics
Branches of a Tree: Taking Derivatives of Programs with Discrete and Branching Randomness in High Energy Physics Open
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 processe…
View article: Machine learning and LHC event generation
Machine learning and LHC event generation Open
First-principle simulations are at the heart of the high-energy physics research program. They link the vast data output of multi-purpose detectors with fundamental theory predictions and interpretation. This review illustrates a wide rang…
View article: Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml
Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml Open
Recurrent neural networks have been shown to be effective architectures for many tasks in high energy physics, and thus have been widely adopted. Their use in low-latency environments has, however, been limited as a result of the difficult…
View article: Differentiable Matrix Elements with MadJax
Differentiable Matrix Elements with MadJax Open
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 incorporatio…
View article: Mass Surrender in Immigration Court
Mass Surrender in Immigration Court Open
View article: Number of spanning trees containing a given forest
Number of spanning trees containing a given forest Open
We consider all spanning trees of a complete simple graph $Γ$ on $n$ vertices that contain a given $m-$forest $F$. We show that the number of such spanning trees, $τ(F)$, doesn't depend on the structure of $F$ and is completely determined …
View article: Report on 2203.07460v1
Report on 2203.07460v1 Open
First-principle simulations are at the heart of the high-energy physics research program.They link the vast data output of multi-purpose detectors with fundamental theory predictions and interpretation.This review illustrates a wide range …
View article: Interpretable Uncertainty Quantification in AI for HEP
Interpretable Uncertainty Quantification in AI for HEP Open
Estimating uncertainty is at the core of performing scientific measurements in HEP: a measurement is not useful without an estimate of its uncertainty. The goal of uncertainty quantification (UQ) is inextricably linked to the question, "ho…
View article: Interpretable Uncertainty Quantification in AI for HEP
Interpretable Uncertainty Quantification in AI for HEP Open
Estimating uncertainty is at the core of performing scientific measurements in HEP: a measurement is not useful without an estimate of its uncertainty. The goal of uncertainty quantification (UQ) is inextricably linked to the question, "ho…
View article: Novel light field imaging device with enhanced light collection for cold atom clouds
Novel light field imaging device with enhanced light collection for cold atom clouds Open
We present a light field imaging system that captures multiple views of an object with a single shot. The system is designed to maximize the total light collection by accepting a larger solid angle of light than a conventional lens with eq…
View article: Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml
Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml Open
Recurrent neural networks have been shown to be effective architectures for many tasks in high energy physics, and thus have been widely adopted. Their use in low-latency environments has, however, been limited as a result of the difficult…
View article: Novel Light Field Imaging Device with Enhanced Light Collection for Cold\n Atom Clouds
Novel Light Field Imaging Device with Enhanced Light Collection for Cold\n Atom Clouds Open
We present a light field imaging system that captures multiple views of an object with a single shot. The system is designed to maximize the total light collection by accepting a larger solid angle of light than a conventional lens with eq…
View article: Jet-images — deep learning edition
Jet-images — deep learning edition Open
Building on the notion of a particle physics detector as a camera and the collimated streams of high energy particles, or jets, it measures as an image, we investigate the potential of machine learning techniques based on deep learning arc…