A neuronal least-action principle for real-time learning in cortical circuits
eLife · 20.12.2024 · 12
DOI: 10.7554/elife.89674 ↗Kurz gefasst
Fachartikel in der Zeitschrift eLife, online erschienen am 20.12.2024. Verfasst von Walter Senn und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Physiology, University of Bern, Bern, Switzerland. Hauptthemen laut MEDLINE: Learning, Models, Neurological, Neuronal Plasticity. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 108 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Union 7th Framework Programme, Swiss National Science Foundation, Horizon Europe.
Abstract (Originalsprache)
One of the most fundamental laws of physics is the principle of least action. Motivated by its predictive power, we introduce a neuronal least-action principle for cortical processing of sensory streams to produce appropriate behavioral outputs in real time. The principle postulates that the voltage dynamics of cortical pyramidal neurons prospectively minimizes the local somato-dendritic mismatch error within individual neurons. For output neurons, the principle implies minimizing an instantaneous behavioral error. For deep network neurons, it implies the prospective firing to overcome integration delays and correct for possible output errors right in time. The neuron-specific errors are extracted in the apical dendrites of pyramidal neurons through a cortical microcircuit that tries to explain away the feedback from the periphery, and correct the trajectory on the fly. Any motor output is in a moving equilibrium with the sensory input and the motor feedback during the ongoing sensory-motor transform. Online synaptic plasticity reduces the somatodendritic mismatch error within each cortical neuron and performs gradient descent on the output cost at any moment in time. The neuronal least-action principle offers an axiomatic framework to derive local neuronal and synaptic laws for global real-time computation and learning in the brain.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Physiology, University of Bern, Bern, Switzerland.
- Kirchhoff-Institute for Physics, Heidelberg University, Heidelberg, Germany.
- European Space Research and Technology Centre, European Space Agency, Noordwijk, Netherlands.
- Insel Data Science Center, University Hospital Bern, Bern, Switzerland.
- Electrical Engineering, Yale University, New Haven, United States.
- MILA, University of Montreal, Montreal, Canada.
ORCID-Profile: Walter Senn, Dominik Dold, Benjamin Ellenberger, Mihai A Petrovici
Förderung
- European Union 7th Framework Programme
- Swiss National Science Foundation
- Horizon Europe
Zitieren
APA
Senn, W., Dold, D., Kungl, A. F., Ellenberger, B., Jordan, J., Bengio, Y., Sacramento, J., & Petrovici, M. A. (2024). A neuronal least-action principle for real-time learning in cortical circuits. eLife, 12. https://doi.org/10.7554/elife.89674
Vancouver
Senn W, Dold D, Kungl AF, Ellenberger B, Jordan J, Bengio Y, et al. A neuronal least-action principle for real-time learning in cortical circuits. eLife. 2024;12. doi:10.7554/elife.89674
BibTeX
@article{Senn2024,
author = {Senn, Walter and Dold, Dominik and Kungl, Akos F and Ellenberger, Benjamin and Jordan, Jakob and Bengio, Yoshua and Sacramento, João and Petrovici, Mihai A},
title = {A neuronal least-action principle for real-time learning in cortical circuits},
journal = {eLife},
year = {2024},
volume = {12},
doi = {10.7554/elife.89674}
}
EUVIMED