Abstract

In recent decades, artificial intelligence has been successively employed in the fields of finance, commerce, and other industries. However, imitating high-level brain functions, such as imagination and inference, pose several challenges as they are relevant to a particular type of noise in a biological neuron network. Probabilistic computing algorithms based on restricted Boltzmann machine and Bayesian inference that use silicon electronics have progressed significantly in terms of mimicking probabilistic inference. However, the quasi-random noise generated from additional circuits or algorithms presents a major challenge for silicon electronics to realize the true stochasticity of biological neuron systems. Artificial neurons based on emerging devices, such as memristors and ferroelectric field-effect transistors with inherent stochasticity can produce uncertain non-linear output spikes, which may be the key to make machine learning closer to the human brain. In this article, we present a comprehensive review of the recent advances in the emerging stochastic artificial neurons (SANs) in terms of probabilistic computing. We briefly introduce the biological neurons, neuron models, and silicon neurons before presenting the detailed working mechanisms of various SANs. Finally, the merits and demerits of silicon-based and emerging neurons are discussed, and the outlook for SANs is presented.

Highlights

  • Chaos is generally undesirable for artificial intelligence architectures, long-term chaotic fluctuations in human brain waves exhibit significant functions in biological neural networks

  • SANs that can effectively mimic the sources of background noise with true stochasticity are essential components in SNNs when used for probabilistic computing

  • Further analyses are required to ensure that the artificial neurons are more bio-mimetic, which warrants dedicated investigations on device dynamics

Read more

Summary

INTRODUCTION

Chaos is generally undesirable for artificial intelligence architectures, long-term chaotic fluctuations in human brain waves exhibit significant functions in biological neural networks. Tuning the input pulse width can generate stochastic outputs with different firing rates Another implementation of a Cu filament-based TS device (Wang et al, 2021) presented LIF neuron behavior by coupling the device with two resistors in series and a capacitor in parallel (Figure 4G). Replacing the tunable resistance with a transistor and adding a thermal noise voltage source [η(t)] (right panel of Figure 11C) renders this neuron circuit sufficiently competent to manipulate the random distribution of threshold voltage of the VO2-based memristor from both thermal and electrical aspects and control the stochastic firing rate rather than the integration rate (Tuma et al, 2016). In comparison with VO2, NbOx is considered a more suitable option for applications at chip level owing to its higher TC (810◦C) (Páez Fajardo et al, 2021)

DISCUSSION
SUMMARY AND OUTLOOK
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call