Mott memristor based stochastic neurons for probabilistic computing.

Abstract

Many studies suggest that probabilistic spiking in biological neural systems is beneficial as it aids learning and provides Bayesian inference-like dynamics. If appropriately utilised, noise and stochasticity in nanoscale devices can benefit neuromorphic systems. In this paper, we build stochastic spiking neurons, a stochastic leaky integrate and fire(LIF) neuron and a probabilistic sigmoid neuron, utilising a Mott memristor's inherent stochastic switching dynamics. We demonstrate that the developed LIF neuron is capable of biological neural dynamics. We leverage these characteristics of the proposed LIF neuron by integrating it into a population-coded spiking neural network, thereby showcasing its ability to implement probabilistic learning and inference. We also utilise the probabilistic sigmoid neuron to implement a Spiking Restricted Boltzmann Machine (sRBM) that achieves a software-comparable accuracy of 85.56\%. Unlike CMOS-based probabilistic neurons, our design does not require any external noise sources. The designed neurons are highly energy efficient and ultra-compact, requiring only three components: a resistor, a capacitor and a memristor device.