Salvato in:
Dettagli Bibliografici
Autori principali: Liao, Qianli, Ziyin, Liu, Gan, Yulu, Cheung, Brian, Harnett, Mark, Poggio, Tomaso
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2412.20018
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915083669471232
author Liao, Qianli
Ziyin, Liu
Gan, Yulu
Cheung, Brian
Harnett, Mark
Poggio, Tomaso
author_facet Liao, Qianli
Ziyin, Liu
Gan, Yulu
Cheung, Brian
Harnett, Mark
Poggio, Tomaso
contents Over the last four decades, the amazing success of deep learning has been driven by the use of Stochastic Gradient Descent (SGD) as the main optimization technique. The default implementation for the computation of the gradient for SGD is backpropagation, which, with its variations, is used to this day in almost all computer implementations. From the perspective of neuroscientists, however, the consensus is that backpropagation is unlikely to be used by the brain. Though several alternatives have been discussed, none is so far supported by experimental evidence. Here we propose a circuit for updating the weights in a network that is biologically plausible, works as well as backpropagation, and leads to verifiable predictions about the anatomy and the physiology of a characteristic motif of four plastic synapses between ascending and descending cortical streams. A key prediction of our proposal is a surprising property of self-assembly of the basic circuit, emerging from initial random connectivity and heterosynaptic plasticity rules.
format Preprint
id arxiv_https___arxiv_org_abs_2412_20018
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Self-Assembly of a Biologically Plausible Learning Circuit
Liao, Qianli
Ziyin, Liu
Gan, Yulu
Cheung, Brian
Harnett, Mark
Poggio, Tomaso
Neural and Evolutionary Computing
Over the last four decades, the amazing success of deep learning has been driven by the use of Stochastic Gradient Descent (SGD) as the main optimization technique. The default implementation for the computation of the gradient for SGD is backpropagation, which, with its variations, is used to this day in almost all computer implementations. From the perspective of neuroscientists, however, the consensus is that backpropagation is unlikely to be used by the brain. Though several alternatives have been discussed, none is so far supported by experimental evidence. Here we propose a circuit for updating the weights in a network that is biologically plausible, works as well as backpropagation, and leads to verifiable predictions about the anatomy and the physiology of a characteristic motif of four plastic synapses between ascending and descending cortical streams. A key prediction of our proposal is a surprising property of self-assembly of the basic circuit, emerging from initial random connectivity and heterosynaptic plasticity rules.
title Self-Assembly of a Biologically Plausible Learning Circuit
topic Neural and Evolutionary Computing
url https://arxiv.org/abs/2412.20018