Learning Fixed Points of Recurrent Neural Networks by Reparameterizing the Network Model
Neural Computation · 2024
DOI: 10.1162/neco_a_01681 ↗Kurz gefasst
Fachartikel in der Zeitschrift Neural Computation, 2024. Verfasst von Vicky Zhu und 1 weiteren Autor:innen. Beteiligte Einrichtung: Babson College, Mathematics, Analytics, Science, and Technology Division, Wellesley, MA 02481, U.S.A. vzhu@babson.edu. Hauptthemen laut MEDLINE: Neural Networks, Computer, Learning. Die Arbeit nennt 55 Literaturangaben. Gefördert von: National Science Foundation.
Abstract (Originalsprache)
In computational neuroscience, recurrent neural networks are widely used to model neural activity and learning. In many studies, fixed points of recurrent neural networks are used to model neural responses to static or slowly changing stimuli, such as visual cortical responses to static visual stimuli. These applications raise the question of how to train the weights in a recurrent neural network to minimize a loss function evaluated on fixed points. In parallel, training fixed points is a central topic in the study of deep equilibrium models in machine learning. A natural approach is to use gradient descent on the Euclidean space of weights. We show that this approach can lead to poor learning performance due in part to singularities that arise in the loss surface. We use a reparameterization of the recurrent network model to derive two alternative learning rules that produce more robust learning dynamics. We demonstrate that these learning rules avoid singularities and learn more effectively than standard gradient descent. The new learning rules can be interpreted as steepest descent and gradient descent, respectively, under a non-Euclidean metric on the space of recurrent weights. Our results question the common, implicit assumption that learning in the brain should be expected to follow the negative Euclidean gradient of synaptic weights.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Babson College, Mathematics, Analytics, Science, and Technology Division, Wellesley, MA 02481, U.S.A. vzhu@babson.edu.
- University of Notre Dame, Department of Applied and Computational Mathematics and Statistics, Notre Dame, IN 46556, U.S.A. Robert.Rosenbaum@nd.edu.
Förderung
- National Science Foundation
Zitieren
APA
Zhu, V., & Rosenbaum, R. (2024). Learning Fixed Points of Recurrent Neural Networks by Reparameterizing the Network Model. Neural Computation. https://doi.org/10.1162/neco_a_01681
Vancouver
Zhu V, Rosenbaum R. Learning Fixed Points of Recurrent Neural Networks by Reparameterizing the Network Model. Neural Computation. 2024. doi:10.1162/neco_a_01681
BibTeX
@article{Zhu2024,
author = {Zhu, Vicky and Rosenbaum, Robert},
title = {Learning Fixed Points of Recurrent Neural Networks by Reparameterizing the Network Model},
journal = {Neural Computation},
year = {2024},
doi = {10.1162/neco_a_01681}
}
EUVIMED