Saved in:
Bibliographic Details
Main Authors: Bahri, Yasaman, Hanin, Boris, Brossollet, Antonin, Erba, Vittorio, Keup, Christian, Pacelli, Rosalba, Simon, James B.
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2309.01592
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909104462626816
author Bahri, Yasaman
Hanin, Boris
Brossollet, Antonin
Erba, Vittorio
Keup, Christian
Pacelli, Rosalba
Simon, James B.
author_facet Bahri, Yasaman
Hanin, Boris
Brossollet, Antonin
Erba, Vittorio
Keup, Christian
Pacelli, Rosalba
Simon, James B.
contents 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 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.
format Preprint
id arxiv_https___arxiv_org_abs_2309_01592
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Les Houches Lectures on Deep Learning at Large & Infinite Width
Bahri, Yasaman
Hanin, Boris
Brossollet, Antonin
Erba, Vittorio
Keup, Christian
Pacelli, Rosalba
Simon, James B.
Machine Learning
Artificial Intelligence
High Energy Physics - Theory
Probability
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 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.
title Les Houches Lectures on Deep Learning at Large & Infinite Width
topic Machine Learning
Artificial Intelligence
High Energy Physics - Theory
Probability
url https://arxiv.org/abs/2309.01592