Les Houches Research Articles

Les Houches guide to reusable ML models in LHC analyses

With the increasing usage of machine-learning in high-energy physics analyses, the publication of the trained models in a reusable form has become a crucial question for analysis preservation and reuse. The complexity of these models creates practical issues for both reporting them accurately and for ensuring the stability of their behaviours in different environments and over extended timescales. In this note we discuss the current state of affairs, highlighting specific practical issues and focusing on the most promising technical and strategic approaches to ensure trustworthy analysis-preservation. This material originated from discussions in the LHC Reinterpretation Forum and the 2023 PhysTeV workshop at Les Houches.

Kernels, data & physics

Abstract Lecture notes from the course given by Professor Julia Kempe at the summer school ‘Statistical physics of Machine Learning’ in Les Houches. The notes discuss the so-called NTK approach to problems in machine learning, which consists of gaining an understanding of generally unsolvable problems by finding a tractable kernel formulation. The notes are mainly focused on practical applications such as data distillation and adversarial robustness, examples of inductive bias are also discussed.

Computational complexity of deep learning: fundamental limitations and empirical phenomena

Abstract This manuscript is the lecture notes of B. Barak’s course in the Les Houches ‘Statistical Physics and Machine Learning’ summer school in 2022. It surveys various proxies for computational hardness in random planted problems, from the low-degree likelihood ratio to statistical query complexity and the Franz–Parisi criterion, as well as the various relationships between those criteria. We also present a few aspects of the study of deep learning, from both a theoretical and empirical point of view.

Les Houches lectures on deep learning at large and infinite width*

Abstract These lectures, presented at the 2022 Les Houches Summer School on Statistical Physics and Machine Learning, focus on the infinite-width limit and large-width regime of deep neural networks. Topics covered include the various statistical and dynamical properties of these networks. In particular, the lecturers discuss properties of random deep neural networks, connections between trained deep neural networks, linear models, kernels and Gaussian processes that arise in the infinite-width limit, and perturbative and non-perturbative treatments of large but finite-width networks, at initialization and after training.

Sparse representations, inference and learning

Abstract In recent years statistical physics has proven to be a valuable tool to probe into large dimensional inference problems such as the ones occurring in machine learning. Statistical physics provides analytical tools to study fundamental limitations in their solutions and proposes algorithms to solve individual instances. In these notes, based on the lectures by Marc Mézard in 2022 at the summer school in Les Houches, we will present a general framework that can be used in a large variety of problems with weak long-range interactions, including the compressed sensing problem, or the problem of learning in a perceptron. We shall see how these problems can be studied at the replica symmetric level, using developments of the cavity methods, both as a theoretical tool and as an algorithm.

Release note: VBFNLO 3.0

Vbfnlo is a flexible parton level Monte Carlo program for the simulation of vector boson fusion (VBF), QCD-induced single and double vector boson production plus two jets, and double and triple vector boson production (plus jet) in hadronic collisions at next-to-leading order (NLO) in the strong coupling constant, as well as Higgs boson plus two and three jet production via gluon fusion at the one-loop level. For the new version – Version 3.0 – several major enhancements have been included. An interface according to the Binoth Les Houches Accord (BLHA) has been added for all VBF and di/tri-boson processes including fully leptonic decays. For all dimension-8 operators affecting vector boson scattering (VBS) processes, a modified T-matrix unitarization procedure has been implemented. Several new production processes have been added, namely the VBS Zγjj\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$Z\\gamma jj$$\\end{document} and γγjj\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma \\gamma jj$$\\end{document} processes at NLO, γγjj\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma \\gamma jj $$\\end{document}, WWj and ZZj production at NLO including the loop-induced gluon-fusion contributions and the gluon-fusion one-loop induced Φjjj\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\varPhi jjj$$\\end{document} (Φ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\varPhi $$\\end{document} is a CP-even or CP-odd scalar boson) process at LO, retaining the full top-mass dependence. Finally, the code has been parallelized using Openmpi.

A brief tutorial on information theory

At the 2023 Les Houches Summer School on Theoretical Biological Physics, several students asked for some background on information theory, and so we added a tutorial to the scheduled lectures. This is largely a transcript of that tutorial, lightly edited. It covers basic definitions and context rather than detailed calculations. We hope to have maintained the informality of the presentation, including exchanges with the students, while still being useful.

Quantum computation with cat qubits

These are the lecture notes from the 2019 Les Houches Summer School on “Quantum Information Machines”. After a brief introduction to quantum error correction and bosonic codes, we focus on the case of cat qubits stabilized by a nonlinear multi-photon driven dissipation process. We argue that such a system can be seen as a self-correcting qubit where bit-flip errors are robustly and exponentially suppressed. Next, we provide some experimental directions to engineer such a multi-photon driven dissipation process with superconducting circuits. Finally, we analyze various logical gates that can be implemented without re-introducing bit-flip errors. This set of bias-preserving gates pave the way towards a hardware-efficient and fault-tolerant quantum processor.

Introduction to quantum error correction and fault tolerance

These lecture notes from the 2019 Les Houches Summer School on “Quantum Information Machines” are intended to provide an introduction to classical and quantum error correction with bits and qubits, and with continuous variable systems (harmonic oscillators). The focus on the latter will be on practical examples that can be realized today or in the near future with a modular architecture based on superconducting electrical circuits and microwave photons. The goal and vision is “hardware-efficient” quantum error correction that does not require exponentially large hardware overhead in order to achieve practical and useful levels of fault tolerance and circuit depth.

SIMP Dark Matter

These notes summarize the zoom-course on Strongly Interacting Massive Particle (SIMP) dark matter given at the Les Houches Summer School in the summer of 2021. An alternative title for this course would be `SIMPS, Cannibals and ELDERs', where our focus is on high-point interactions amongst dark matter particles. The spirit of the course is to give students a taste of these exciting developments in dark matter, while primarily teaching tricks and methods that are tough to captivate when reading textbooks or papers. These notes are written in similar spirit: I will put emphasis on how quantities scale and how to perform back of the envelope estimates, which should serve to help you when you develop the next great dark matter idea.

Dark Matter direct detection of classical WIMPs

One of the highest priorities in particle physics today is the identification of the constituents of dark matter. This manuscript is a supplement to pedagogical lectures given at the 2021 Les Houches Summer School on Dark Matter. The lectures cover topics related to the direct detection of Weakly Interaction Massive Particle (WIMP) dark matter, including the distribution of dark matter, nuclear scattering, backgrounds, planning and designing of experiments, and a sampling of planned and ongoing experiments.

Les Houches Lectures on Indirect Detection of Dark Matter

These lectures, presented at the 2021 Les Houches Summer School on Dark Matter, provide an introduction to key methods and tools of indirect dark matter searches, as well as a status report on the field circa summer 2021. Topics covered include the possible effects of energy injection from dark matter on the early universe, methods to calculate both the expected energy distribution and spatial distribution of particles produced by dark matter interactions, an outline of theoretical models that predict diverse signals in indirect detection, and a discussion of current constraints and some claimed anomalies. These notes are intended as an introduction to indirect dark matter searches for graduate students, focusing primarily on intuition-building estimates and useful concepts and tools.

Dark atoms and composite dark matter

I selectively review the theoretical properties and observational limits pertaining to dark atoms, as well as composite dark matter candidates bound by a confining gauge interaction: dark glueballs, glueballinos, mesons and baryons. Emphasis is given to cosmological, direct and indirect detection constraints. Lectures given at Les Houches Summer School 2021: Dark Matter.

An introduction to axions and their detection

In these notes I try to introduce the reader to the topic of axions: their theoretical motivation and expected phenomenology, their role in astrophysics and as a dark matter candidate, and the experimental techniques to detect them. Special emphasis is made in this last point, for which a relatively updated review of worldwide efforts and future prospects is made. The material is intended as an introduction to the topic, and it was prepared as lecture notes for Les Houches summer school 2021. Abundant references are included to direct the reader to deeper insight on the different aspects of axion physics.

Introduction to quantum non-reciprocal interactions: from non-Hermitian Hamiltonians to quantum master equations and quantum feedforward schemes

These lecture notes from the 2019 Les Houches Summer School on Quantum Information Machines provides a pedagogical introduction to the theory of non-reciprocal quantum interactions and devices. The goal is connect various approaches and concepts, including Hamiltonians encoding synthetic gauge fields, scattering descriptions, quantum master equations, and non-Hermitian Hamiltonians. The importance of having both non-trivial synthetic gauge fields and dissipation for obtaining non-reciprocal interactions is stressed. Connections to broader topics such as quantum reservoir engineering and the quantum theory of continuous-measurement based feedforward are also discussed.

Sub-GeV dark matter models and direct detection

Lecture notes for Les Houches Summer School 2021: Dark Matter. These lectures give a brief introduction to sub-GeV dark matter models, and then reviews theory and approaches to direct detection of sub-GeV dark matter.

Quantum enhanced metrology in the search for fundamental physical phenomena

These notes summarize lectures given at the 2019 Les Houches summer school on Quantum Information Machines. They describe and review an application of quantum metrology concepts to searches for ultralight dark matter. In particular, for ultralight dark matter that couples as a weak classical force to a laboratory harmonic oscillator, quantum squeezing benefits experiments in which the mass of the dark matter particle is unknown. This benefit is present even if the oscillatory dark matter signal is much more coherent than the harmonic oscillator that it couples to, as is the case for microwave frequency searches for dark matter axion particles.

The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider

We describe the outcome of a data challenge conducted as part of the Dark Machines (https://www.darkmachines.org) initiative and the Les Houches 2019 workshop on Physics at TeV colliders. The challenged aims to detect signals of new physics at the Large Hadron Collider (LHC) using unsupervised machine learning algorithms. First, we propose how an anomaly score could be implemented to define model-independent signal regions in LHC searches. We define and describe a large benchmark dataset, consisting of >1 billion simulated LHC events corresponding to 10\, fb^{-1}10fb−1 of proton-proton collisions at a center-of-mass energy of 13 TeV. We then review a wide range of anomaly detection and density estimation algorithms, developed in the context of the data challenge, and we measure their performance in a set of realistic analysis environments. We draw a number of useful conclusions that will aid the development of unsupervised new physics searches during the third run of the LHC, and provide our benchmark dataset for future studies at https://www.phenoMLdata.org. Code to reproduce the analysis is provided at https://github.com/bostdiek/DarkMachines-UnsupervisedChallenge.

Neutrino-nucleon DIS from holographic QCD: PDFs of sea and valence quarks, form factors, and structure functions of the proton

We discuss unpolarized neutrino- and anti-neutrino-nucleon deep inelastic scattering using a chiral doublet of baryonic sources with explicit symmetry breaking, in a slice of ${\mathrm{AdS}}_{5}$ with both a hard and soft wall. We explicitly derive the direct and transition form factors for the vector and axial-vector currents for the holographic dual of a proton and neutron. We use them to derive the $s$-channel structure functions for neutrino and antineutrino scattering on a proton and neutron in bulk. The $t$-channel contributions stemming from the Pomeron and Reggeon exchanges are also evaluated explicitly. The pertinent even and odd structure functions in the limit of large and small parton momentum fraction $x$ are given. The results allow for the extraction of the nonperturbative parton distribution functions carried by the sea and valence quarks both at large-$x$ and small-$x$ regimes. Our holographic parton distribution function (PDF) sets compare well with the Les Houches accord PDF (LHAPDF) and the coordinated theoretical-experiment project on QCD (CTEQ) PDF sets in the large-$x$ and small-$x$ regimes in the intermediate range of ${Q}^{2}<10\text{ }\text{ }{\mathrm{GeV}}^{2}$.

Updates to the one-loop provider NLOX

We describe the 1.2 update to NLOX, a computer program for calculations in high-energy particle physics. New features since the 1.0 release and other changes are described, along with usage documentation.