Saved in:
Bibliographic Details
Main Authors: Kubo, Yoshimasa, Modi, Suhani Pragnesh, Patel, Smit
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.03402
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908865028685824
author Kubo, Yoshimasa
Modi, Suhani Pragnesh
Patel, Smit
author_facet Kubo, Yoshimasa
Modi, Suhani Pragnesh
Patel, Smit
contents Equilibrium propagation (EP) is a biologically plausible alternative to backpropagation for training neural networks. However, existing EP models use a uniform scalar time step dt, which corresponds biologically to a membrane time constant that is heterogeneous across neurons. Here, we introduce heterogeneous time steps (HTS) for EP by assigning neuron-specific time constants drawn from biologically motivated distributions. We show that HTS improves training stability while maintaining competitive task performance. These results suggest that incorporating heterogeneous temporal dynamics enhances both the biological realism and robustness of equilibrium propagation.
format Preprint
id arxiv_https___arxiv_org_abs_2603_03402
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Heterogeneous Time Constants Improve Stability in Equilibrium Propagation
Kubo, Yoshimasa
Modi, Suhani Pragnesh
Patel, Smit
Machine Learning
Artificial Intelligence
Equilibrium propagation (EP) is a biologically plausible alternative to backpropagation for training neural networks. However, existing EP models use a uniform scalar time step dt, which corresponds biologically to a membrane time constant that is heterogeneous across neurons. Here, we introduce heterogeneous time steps (HTS) for EP by assigning neuron-specific time constants drawn from biologically motivated distributions. We show that HTS improves training stability while maintaining competitive task performance. These results suggest that incorporating heterogeneous temporal dynamics enhances both the biological realism and robustness of equilibrium propagation.
title Heterogeneous Time Constants Improve Stability in Equilibrium Propagation
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2603.03402