Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arxiv logo > stat > arXiv:2211.11567

Help | Advanced Search

arXiv logo
Cornell University Logo

quick links

  • Login
  • Help Pages
  • About

Statistics > Machine Learning

arXiv:2211.11567 (stat)
[Submitted on 21 Nov 2022 (v1), last revised 26 May 2023 (this version, v2)]

Title:Neural networks trained with SGD learn distributions of increasing complexity

Authors:Maria Refinetti, Alessandro Ingrosso, Sebastian Goldt
View a PDF of the paper titled Neural networks trained with SGD learn distributions of increasing complexity, by Maria Refinetti and Alessandro Ingrosso and Sebastian Goldt
View PDF
Abstract:The ability of deep neural networks to generalise well even when they interpolate their training data has been explained using various "simplicity biases". These theories postulate that neural networks avoid overfitting by first learning simple functions, say a linear classifier, before learning more complex, non-linear functions. Meanwhile, data structure is also recognised as a key ingredient for good generalisation, yet its role in simplicity biases is not yet understood. Here, we show that neural networks trained using stochastic gradient descent initially classify their inputs using lower-order input statistics, like mean and covariance, and exploit higher-order statistics only later during training. We first demonstrate this distributional simplicity bias (DSB) in a solvable model of a neural network trained on synthetic data. We empirically demonstrate DSB in a range of deep convolutional networks and visual transformers trained on CIFAR10, and show that it even holds in networks pre-trained on ImageNet. We discuss the relation of DSB to other simplicity biases and consider its implications for the principle of Gaussian universality in learning.
Comments: Source code available at this https URL
Subjects: Machine Learning (stat.ML); Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech); Machine Learning (cs.LG)
Cite as: arXiv:2211.11567 [stat.ML]
  (or arXiv:2211.11567v2 [stat.ML] for this version)
  https://doi.org/10.48550/arXiv.2211.11567
arXiv-issued DOI via DataCite
Journal reference: ICML 2023

Submission history

From: Sebastian Goldt [view email]
[v1] Mon, 21 Nov 2022 15:27:22 UTC (4,863 KB)
[v2] Fri, 26 May 2023 13:11:17 UTC (8,892 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Neural networks trained with SGD learn distributions of increasing complexity, by Maria Refinetti and Alessandro Ingrosso and Sebastian Goldt
  • View PDF
  • TeX Source
  • Other Formats
view license
Current browse context:
cs
< prev   |   next >
new | recent | 2022-11
Change to browse by:
cond-mat
cond-mat.dis-nn
cond-mat.stat-mech
cs.LG
stat
stat.ML

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar
a export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
  • Author
  • Venue
  • Institution
  • Topic

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
  • About
  • Help
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status
    Get status notifications via email or slack