EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Approximating Nonlinear Functions With Latent Boundaries in Low-Rank Excitatory-Inhibitory Spiking Networks

William F Podlaski, Christian K Machens

Neural Computation · 2024

DOI: 10.1162/neco_a_01658 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Neural Computation, 2024. Verfasst von William F Podlaski und 1 weiteren Autor:innen. Beteiligte Einrichtung: Champalimaud Neuroscience Programme, Champalimaud Foundation, 1400-038 Lisbon, Portugal william.podlaski@research.fchampalimaud.org. Hauptthemen laut MEDLINE: Action Potentials, Nonlinear Dynamics, Neurons, Neural Networks, Computer. Die Arbeit wurde bisher 21-mal zitiert (Crossref) und nennt 106 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Fundação para a Ciência e a Tecnologia, Simons Foundation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Deep feedforward and recurrent neural networks have become successful functional models of the brain, but they neglect obvious biological details such as spikes and Dale’s law. Here we argue that these details are crucial in order to understand how real neural circuits operate. Towards this aim, we put forth a new framework for spike-based computation in low-rank excitatory-inhibitory spiking networks. By considering populations with rank-1 connectivity, we cast each neuron’s spiking threshold as a boundary in a low-dimensional input-output space. We then show how the combined thresholds of a population of inhibitory neurons form a stable boundary in this space, and those of a population of excitatory neurons form an unstable boundary. Combining the two boundaries results in a rank-2 excitatory-inhibitory (EI) network with inhibition-stabilized dynamics at the intersection of the two boundaries. The computation of the resulting networks can be understood as the difference of two convex functions and is thereby capable of approximating arbitrary non-linear input-output mappings. We demonstrate several properties of these networks, including noise suppression and amplification, irregular activity and synaptic balance, as well as how they relate to rate network dynamics in the limit that the boundary becomes soft. Finally, while our work focuses on small networks (5-50 neurons), we discuss potential avenues for scaling up to much larger networks. Overall, our work proposes a new perspective on spiking networks that may serve as a starting point for a mechanistic understanding of biological spike-based computation.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Action PotentialsNonlinear DynamicsNeuronsNeural Networks, ComputerNeural InhibitionModels, NeurologicalHumansAnimalsNerve NetSynapsesComputer Simulation

Autor:innen und Institutionen

  • Champalimaud Neuroscience Programme, Champalimaud Foundation, 1400-038 Lisbon, Portugal william.podlaski@research.fchampalimaud.org.
  • Champalimaud Neuroscience Programme, Champalimaud Foundation, 1400-038 Lisbon, Portugal christian.machens@neuro.fchampalimaud.org.

Förderung

  • Fundação para a Ciência e a Tecnologia
  • Simons Foundation

Zitieren

APA

Podlaski, W. F., & Machens, C. K. (2024). Approximating Nonlinear Functions With Latent Boundaries in Low-Rank Excitatory-Inhibitory Spiking Networks. Neural Computation. https://doi.org/10.1162/neco_a_01658

Vancouver

Podlaski WF, Machens CK. Approximating Nonlinear Functions With Latent Boundaries in Low-Rank Excitatory-Inhibitory Spiking Networks. Neural Computation. 2024. doi:10.1162/neco_a_01658

BibTeX
@article{Podlaski2024,
  author = {Podlaski, William F and Machens, Christian K},
  title = {Approximating Nonlinear Functions With Latent Boundaries in Low-Rank Excitatory-Inhibitory Spiking Networks},
  journal = {Neural Computation},
  year = {2024},
  doi = {10.1162/neco_a_01658}
}
Kostenlos starten →