Random weights of DNNs and emergence of fixed points

DOI10.3934/NHM.2026007MaRDI QIDQ7249453zbMATH OpenFDO

Authors Leonid Berlyand, Oleksii Krupchytskyi, V. A. Slavin

Publication date 15 May 2026

Published in Networks and Heterogeneous Media (Search for Journal in Brave)

Full work available at URL https://doi.org/10.3934/NHM.2026007

zbMATH Keywords

autoencoder deep neural networkfixed pointsrandom initializationstability and basin of attraction

Mathematics Subject Classification ID

Random matrices (probabilistic aspects) (60B20) Artificial neural networks and deep learning (68T07)

Cites work

• An Introduction to Variational Autoencoders
• Deep learning
• Deep Neural Networks with Random Gaussian Weights: A Universal Classification Strategy?
• Deformed Cauchy random matrix ensembles and large \(N\) phase transitions
• DISTRIBUTION OF EIGENVALUES FOR SOME SETS OF RANDOM MATRICES
• Enhancing accuracy in deep learning using random matrix theory
• Fixed Points of Competitive Threshold-Linear Networks
• scientific article; zbMATH DE number 7415108 (Why is no real title available?)
• Mathematics of deep learning. An introduction
• Neural networks and physical systems with emergent collective computational abilities
• Neurons with graded response have collective computational properties like those of two-state neurons
• On random matrices arising in deep neural networks: General I.I.D. case
• On the converse of Banach "fixed-point principle"
• On universality of local edge regime for the deformed Gaussian unitary ensemble
• The Banach fixed point theorem: selected topics from its hundred-year history

This page was built for publication: Random weights of DNNs and emergence of fixed points

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7249453)