Saved in:
Bibliographic Details
Main Authors: Jeon, Hong Jun, Van Roy, Benjamin
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.01456
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910509500989440
author Jeon, Hong Jun
Van Roy, Benjamin
author_facet Jeon, Hong Jun
Van Roy, Benjamin
contents Neural scaling laws aim to characterize how out-of-sample error behaves as a function of model and training dataset size. Such scaling laws guide allocation of a computational resources between model and data processing to minimize error. However, existing theoretical support for neural scaling laws lacks rigor and clarity, entangling the roles of information and optimization. In this work, we develop rigorous information-theoretic foundations for neural scaling laws. This allows us to characterize scaling laws for data generated by a two-layer neural network of infinite width. We observe that the optimal relation between data and model size is linear, up to logarithmic factors, corroborating large-scale empirical investigations. Concise yet general results of the kind we establish may bring clarity to this topic and inform future investigations.
format Preprint
id arxiv_https___arxiv_org_abs_2407_01456
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Information-Theoretic Foundations for Neural Scaling Laws
Jeon, Hong Jun
Van Roy, Benjamin
Machine Learning
Artificial Intelligence
Neural scaling laws aim to characterize how out-of-sample error behaves as a function of model and training dataset size. Such scaling laws guide allocation of a computational resources between model and data processing to minimize error. However, existing theoretical support for neural scaling laws lacks rigor and clarity, entangling the roles of information and optimization. In this work, we develop rigorous information-theoretic foundations for neural scaling laws. This allows us to characterize scaling laws for data generated by a two-layer neural network of infinite width. We observe that the optimal relation between data and model size is linear, up to logarithmic factors, corroborating large-scale empirical investigations. Concise yet general results of the kind we establish may bring clarity to this topic and inform future investigations.
title Information-Theoretic Foundations for Neural Scaling Laws
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2407.01456