Format

Send to

Choose Destination
Proc Natl Acad Sci U S A. 2019 Nov 5;116(45):22811-22820. doi: 10.1073/pnas.1905926116. Epub 2019 Oct 21.

Simple framework for constructing functional spiking recurrent neural networks.

Kim R1,2,3, Li Y4, Sejnowski TJ1,5,6.

Author information

1
Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037; rkim@salk.edu terry@salk.edu.
2
Neurosciences Graduate Program, University of California San Diego, La Jolla, CA 92093.
3
Medical Scientist Training Program, University of California San Diego, La Jolla, CA 92093.
4
Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA 92037.
5
Institute for Neural Computation, University of California San Diego, La Jolla, CA 92093.
6
Division of Biological Sciences, University of California San Diego, La Jolla, CA 92093.

Abstract

Cortical microcircuits exhibit complex recurrent architectures that possess dynamically rich properties. The neurons that make up these microcircuits communicate mainly via discrete spikes, and it is not clear how spikes give rise to dynamics that can be used to perform computationally challenging tasks. In contrast, continuous models of rate-coding neurons can be trained to perform complex tasks. Here, we present a simple framework to construct biologically realistic spiking recurrent neural networks (RNNs) capable of learning a wide range of tasks. Our framework involves training a continuous-variable rate RNN with important biophysical constraints and transferring the learned dynamics and constraints to a spiking RNN in a one-to-one manner. The proposed framework introduces only 1 additional parameter to establish the equivalence between rate and spiking RNN models. We also study other model parameters related to the rate and spiking networks to optimize the one-to-one mapping. By establishing a close relationship between rate and spiking models, we demonstrate that spiking RNNs could be constructed to achieve similar performance as their counterpart continuous rate networks.

KEYWORDS:

rate neural networks; recurrent neural networks; spiking neural networks

PMID:
31636215
PMCID:
PMC6842655
[Available on 2020-04-21]
DOI:
10.1073/pnas.1905926116

Supplemental Content

Full text links

Icon for HighWire
Loading ...
Support Center