Work

Backpropagation through space, time and the brain

Year: 2025

Type: article

Abstract: How physical neuronal networks, bound by spatio-temporal locality constraints, can perform efficient credit assignment, remains an intriguing question. Both backward- and forward-propagation algorithms rely on assumptions that violate this locality in various ways. We introduce Generalized Latent Equilibrium (GLE), a framework for fully local spatio-temporal credit assignment in physical, dynamical neuronal networks. From an energy based on neuron-local mismatches, we derive neuronal dynamics via stationarity and parameter dynamics as gradient descent. The result is an online approximation of backpropagation through space and time in deep networks of cortical microcircuits with continuously active, local synaptic plasticity. GLE exploits dendritic morphology to enable complex information storage and processing in single neurons, as well as their ability to react in anticipation of their future input. This "prospective coding" enables the computation of spatio-temporal convolutions in the forward direction and the approximation of adjoint variables in the backward stream.

Source: Zurich Open Repository and Archive (University of Zurich)

DOI: 10.5167/uzh-292308

Authors: Ellenberger, Benjamin, Haider, Paul, Benitez, Federico, Jordan, Jakob, Max, Kevin, Jaras, Ismael, Kriener, Laura, Petrovici, Mihai A.

Language: English

Topic: Neural dynamics and brain function

Subfield: Cognitive Neuroscience

Field: Neuroscience

Domain: Life Sciences

Open Access status: green

OpenAlex ID: W7130859486

FWCI:

Cites:

Cited by:

Related to: