Saved in:
Bibliographic Details
Main Authors: Gamba, Matteo, Azizpour, Hossein, Björkman, Mårten
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2301.12309
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916857337872384
author Gamba, Matteo
Azizpour, Hossein
Björkman, Mårten
author_facet Gamba, Matteo
Azizpour, Hossein
Björkman, Mårten
contents Existing bounds on the generalization error of deep networks assume some form of smooth or bounded dependence on the input variable, falling short of investigating the mechanisms controlling such factors in practice. In this work, we present an extensive experimental study of the empirical Lipschitz constant of deep networks undergoing double descent, and highlight non-monotonic trends strongly correlating with the test error. Building a connection between parameter-space and input-space gradients for SGD around a critical point, we isolate two important factors -- namely loss landscape curvature and distance of parameters from initialization -- respectively controlling optimization dynamics around a critical point and bounding model function complexity, even beyond the training data. Our study presents novels insights on implicit regularization via overparameterization, and effective model complexity for networks trained in practice.
format Preprint
id arxiv_https___arxiv_org_abs_2301_12309
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle On the Lipschitz Constant of Deep Networks and Double Descent
Gamba, Matteo
Azizpour, Hossein
Björkman, Mårten
Machine Learning
Existing bounds on the generalization error of deep networks assume some form of smooth or bounded dependence on the input variable, falling short of investigating the mechanisms controlling such factors in practice. In this work, we present an extensive experimental study of the empirical Lipschitz constant of deep networks undergoing double descent, and highlight non-monotonic trends strongly correlating with the test error. Building a connection between parameter-space and input-space gradients for SGD around a critical point, we isolate two important factors -- namely loss landscape curvature and distance of parameters from initialization -- respectively controlling optimization dynamics around a critical point and bounding model function complexity, even beyond the training data. Our study presents novels insights on implicit regularization via overparameterization, and effective model complexity for networks trained in practice.
title On the Lipschitz Constant of Deep Networks and Double Descent
topic Machine Learning
url https://arxiv.org/abs/2301.12309