Abstract:The dynamics of neuronal systems are characterized by hallmark features such as oscillations and synchrony. However, it has remained unclear whether these characteristics are epiphenomena or are exploited for computation. Due to the challenge of specifically interfering with oscillatory network dynamics in neuronal systems, we simulated recurrent networks (RNNs) of damped harmonic oscillators in which oscillatory activity is enforced in each node, a choice well-supported by experimental findings. When trained on standard pattern recognition tasks, these harmonic oscillator networks (HORNs) outperformed non-oscillatory architectures with respect to learning speed, noise tolerance, and parameter efficiency. HORNs also reproduced a substantial number of characteristic features of neuronal systems such as the cerebral cortex and the hippocampus. In trained HORNs, stimulus-induced interference patterns holistically represent the result of comparing sensory evidence with priors stored in recurrent connection weights, and learning-induced weight changes are compatible with Hebbian principles. Implementing additional features characteristic of natural networks, such as heterogeneous oscillation frequencies, inhomogeneous conduction delays, and network modularity, further enhanced HORN performance without requiring additional parameters. Taken together, our model allows us to give plausible a posteriori explanations for features of natural networks whose computational role has remained elusive. We conclude that neuronal systems are likely to exploit the unique dynamics of recurrent oscillator networks whose computational superiority critically depends on the oscillatory patterning of their nodal dynamics. Implementing the proposed computational principles in analog hardware is expected to enable the design of highly energy-efficient and self-adapting devices that could ideally complement existing digital technologies.

Hardware implementation of the ring generator with tunable frequency based on electronic neurons

Hardware implementation of the ring generator with tunable frequency based on electronic neurons

Stochastic Memristive Interface between Electronic FitzHugh-Nagumo Neurons

Chaos in A Neuron Model with Adaptive Feedback Synapse: Electronic Circuit Design

Design of a New Neuro-Generator with a Neuronal Module to Produce Pseudorandom and Perfectly Pseudorandom Sequences

Noise driven neuromorphic tuned amplifier

Transient dynamics in electronic neuron-like circuits in application to modeling epileptic seizures

Birhythmic Analog Circuit Maze: A Nonlinear Neurostimulation Testbed

Gamma Rhythm Analysis and Simulation Using Neuron Models

Discrete Modeling of Multi-Transmitter Neural Networks with Neuron Competition

A neuromimetic realization of hippocampal CA1 for theta wave generation

Single‐Transistor Neuron with Excitatory–Inhibitory Spatiotemporal Dynamics Applied for Neuronal Oscillations

High-order sensory processing nanocircuit based on coupled VO2 oscillators

Hardware-Friendly Implementation of Physical Reservoir Computing with CMOS-based Time-domain Analog Spiking Neurons

Stochastic Memristive Interface for Neural Signal Processing

Compact hybrid type electronic neuron and computational model of its dynamics

The functional role of oscillatory dynamics of neocortical circuits: a computational perspective

Dynamic Behavior of Coupling Heterogeneous Neurons and Its Hardware Implementation

Bifurcation and Oscillations of a Multi-ring Coupling Neural Network with Discrete Delays

Memristor-based model of neuronal excitability and synaptic potentiation

Construction and Application of a Neuromorphic Circuit With Excitatory and Inhibitory Post-Synaptic Conduction Channels Implemented Using Dual-Gate Thin-Film Transistors