English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Artificial neural networks can harness stochasticity in multiple ways toenable a vast class of computationally powerful models. Electronicimplementation of such stochastic networks is currently limited to addition ofalgorithmic noise to digital machines which is inherently inefficient; albeitrecent efforts to harness physical noise in devices for stochasticity haveshown promise. To succeed in fabricating electronic neuromorphic networks weneed experimental evidence of devices with measurable and controllablestochasticity which is complemented with the development of reliablestatistical models of such observed stochasticity. Current research literaturehas sparse evidence of the former and a complete lack of the latter. Thismotivates the current article where we demonstrate a stochastic neuron using aninsulator-metal-transition (IMT) device, based on electrically inducedphase-transition, in series with a tunable resistance. We show that an IMTneuron has dynamics similar to a piecewise linear FitzHugh-Nagumo (FHN) neuronand incorporates all characteristics of a spiking neuron in the devicephenomena. We experimentally demonstrate spontaneous stochastic spiking alongwith electrically controllable firing probabilities using Vanadium Dioxide(VO$_2$) based IMT neurons which show a sigmoid-like transfer function. Thestochastic spiking is explained by two noise sources - thermal noise andthreshold fluctuations, which act as precursors of bifurcation. As such, theIMT neuron is modeled as an Ornstein-Uhlenbeck (OU) process with a fluctuatingboundary resulting in transfer curves that closely match experiments. As one ofthe first comprehensive studies of a stochastic neuron hardware and itsstatistical properties, this article would enable efficient implementation of alarge class of neuro-mimetic networks and algorithms.

Stochastic IMT (insulator-metal-transition) neurons: An interplay of thermal and threshold noise at bifurcation