Saved in:
Bibliographic Details
Main Authors: Fayyazi, Ryan, Weilbach, Christian, Wood, Frank
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.05494
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910518503014400
author Fayyazi, Ryan
Weilbach, Christian
Wood, Frank
author_facet Fayyazi, Ryan
Weilbach, Christian
Wood, Frank
contents Inter-neuron communication delays are ubiquitous in physically realized neural networks such as biological neural circuits and neuromorphic hardware. These delays have significant and often disruptive consequences on network dynamics during training and inference. It is therefore essential that communication delays be accounted for, both in computational models of biological neural networks and in large-scale neuromorphic systems. Nonetheless, communication delays have yet to be comprehensively addressed in either domain. In this paper, we first show that delays prevent state-of-the-art continuous-time neural networks called Latent Equilibrium (LE) networks from learning even simple tasks despite significant overparameterization. We then propose to compensate for communication delays by predicting future signals based on currently available ones. This conceptually straightforward approach, which we call prospective messaging (PM), uses only neuron-local information, and is flexible in terms of memory and computation requirements. We demonstrate that incorporating PM into delayed LE networks prevents reaction lags, and facilitates successful learning on Fourier synthesis and autoregressive video prediction tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2407_05494
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Prospective Messaging: Learning in Networks with Communication Delays
Fayyazi, Ryan
Weilbach, Christian
Wood, Frank
Machine Learning
Neural and Evolutionary Computing
Inter-neuron communication delays are ubiquitous in physically realized neural networks such as biological neural circuits and neuromorphic hardware. These delays have significant and often disruptive consequences on network dynamics during training and inference. It is therefore essential that communication delays be accounted for, both in computational models of biological neural networks and in large-scale neuromorphic systems. Nonetheless, communication delays have yet to be comprehensively addressed in either domain. In this paper, we first show that delays prevent state-of-the-art continuous-time neural networks called Latent Equilibrium (LE) networks from learning even simple tasks despite significant overparameterization. We then propose to compensate for communication delays by predicting future signals based on currently available ones. This conceptually straightforward approach, which we call prospective messaging (PM), uses only neuron-local information, and is flexible in terms of memory and computation requirements. We demonstrate that incorporating PM into delayed LE networks prevents reaction lags, and facilitates successful learning on Fourier synthesis and autoregressive video prediction tasks.
title Prospective Messaging: Learning in Networks with Communication Delays
topic Machine Learning
Neural and Evolutionary Computing
url https://arxiv.org/abs/2407.05494