Saved in:
Bibliographic Details
Main Authors: Guillin, Arnaud, Nectoux, Boris, Stos, Paul
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.01842
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914363273641984
author Guillin, Arnaud
Nectoux, Boris
Stos, Paul
author_facet Guillin, Arnaud
Nectoux, Boris
Stos, Paul
contents We quantify, uniformly over time and with high probability, the discrepancy between the predictions of a two-layer neural network trained by stochastic gradient descent (SGD) and their mean-field limit, for quadratic loss and ridge regularization. As a key ingredient, we establish T p transportation inequalities (p $\in$ {1, 2}) for the law of the SGD parameters, with explicit constants independent of the iteration index. We then prove uniform-in-time concentration of the empirical parameter measure around its mean-field limit in the Wasserstein distance W 1 , and we translate these bounds into prediction-error estimates against a fixed test function $Φ$. We also derive analogous concentration bounds in the sliced-Wasserstein distance SW 1 , leading to dimension-free rates.
format Preprint
id arxiv_https___arxiv_org_abs_2603_01842
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Uniform-in-time concentration in two-layer neural networks via transportation inequalities
Guillin, Arnaud
Nectoux, Boris
Stos, Paul
Neural and Evolutionary Computing
Probability
We quantify, uniformly over time and with high probability, the discrepancy between the predictions of a two-layer neural network trained by stochastic gradient descent (SGD) and their mean-field limit, for quadratic loss and ridge regularization. As a key ingredient, we establish T p transportation inequalities (p $\in$ {1, 2}) for the law of the SGD parameters, with explicit constants independent of the iteration index. We then prove uniform-in-time concentration of the empirical parameter measure around its mean-field limit in the Wasserstein distance W 1 , and we translate these bounds into prediction-error estimates against a fixed test function $Φ$. We also derive analogous concentration bounds in the sliced-Wasserstein distance SW 1 , leading to dimension-free rates.
title Uniform-in-time concentration in two-layer neural networks via transportation inequalities
topic Neural and Evolutionary Computing
Probability
url https://arxiv.org/abs/2603.01842