Critical Points of Random Neural Networks

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Di Lillo, Simmaco
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908577760804864
author Di Lillo, Simmaco
author_facet Di Lillo, Simmaco
contents This work investigates the expected number of critical points of random neural networks with different activation functions as the depth increases in the infinite-width limit. Under suitable regularity conditions, we derive precise asymptotic formulas for the expected number of critical points of fixed index and those exceeding a given threshold. Our analysis reveals three distinct regimes depending on the value of the first derivative of the covariance evaluated at 1: the expected number of critical points may converge, grow polynomially, or grow exponentially with depth. The theoretical predictions are supported by numerical experiments. Moreover, we provide numerical evidence suggesting that, when the regularity condition is not satisfied (e.g. for neural networks with ReLU as activation function), the number of critical points increases as the map resolution increases, indicating a potential divergence in the number of critical points.
format Preprint
id arxiv_https___arxiv_org_abs_2505_17000
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Critical Points of Random Neural Networks
Di Lillo, Simmaco
Machine Learning
Probability
60G60, 62B10, 62M45
This work investigates the expected number of critical points of random neural networks with different activation functions as the depth increases in the infinite-width limit. Under suitable regularity conditions, we derive precise asymptotic formulas for the expected number of critical points of fixed index and those exceeding a given threshold. Our analysis reveals three distinct regimes depending on the value of the first derivative of the covariance evaluated at 1: the expected number of critical points may converge, grow polynomially, or grow exponentially with depth. The theoretical predictions are supported by numerical experiments. Moreover, we provide numerical evidence suggesting that, when the regularity condition is not satisfied (e.g. for neural networks with ReLU as activation function), the number of critical points increases as the map resolution increases, indicating a potential divergence in the number of critical points.
title Critical Points of Random Neural Networks
topic Machine Learning
Probability
60G60, 62B10, 62M45
url https://arxiv.org/abs/2505.17000