Artificial Neural Microcircuits as Building Blocks: Concept and Challenges

Andrew Walter; Shimeng Wu; Andy Tyrrell; Liam McDaid; Malachy McElholm; Nidhin Thandassery Sumithran; Jim Harkin; Martin Albrecht Trefzer

doi:10.48550/arXiv.2403.16327

Artificial Neural Microcircuits as Building Blocks: Concept and Challenges

Andrew Walter, Shimeng Wu, Andy Tyrrell, Liam McDaid, Malachy McElholm, Nidhin Thandassery Sumithran, Jim Harkin, Martin Albrecht Trefzer

Electronic Engineering

Research output: Working paper › Preprint

Abstract

Artificial Neural Networks (ANNs) are one of the most widely employed forms of bio-inspired computation. However the current trend is for ANNs to be structurally homogeneous. Furthermore, this structural homogeneity requires the application of complex training and learning tools that produce application specific ANNs, susceptible to pitfalls such as overfitting. In this paper, an new approach is explored, inspired by the role played in biology by Neural Microcircuits, the so called ``fundamental processing elements'' of organic nervous systems. How large neural networks, particularly Spiking Neural Networks (SNNs) can be assembled using Artificial Neural Microcircuits (ANMs), intended as off-the-shelf components, is articulated; the results of initial work to produce a catalogue of such Microcircuits though the use of Novelty Search is shown; followed by efforts to expand upon this initial work, including a discussion of challenges uncovered during these efforts and explorations of methods by which they might be overcome.

Original language	English
Publisher	arXiv
DOIs	https://doi.org/10.48550/arXiv.2403.16327
Publication status	Published - 24 Mar 2024

Keywords

artificial neural networks
neural microcircuits
neuromorphic systems
fault tolerance
electronic systems

Access to Document

10.48550/arXiv.2403.16327

Cite this

@techreport{f80eeeed31b34f88aaf7295d2df2d201,

title = "Artificial Neural Microcircuits as Building Blocks: Concept and Challenges",

abstract = "Artificial Neural Networks (ANNs) are one of the most widely employed forms of bio-inspired computation. However the current trend is for ANNs to be structurally homogeneous. Furthermore, this structural homogeneity requires the application of complex training and learning tools that produce application specific ANNs, susceptible to pitfalls such as overfitting. In this paper, an new approach is explored, inspired by the role played in biology by Neural Microcircuits, the so called ``fundamental processing elements'' of organic nervous systems. How large neural networks, particularly Spiking Neural Networks (SNNs) can be assembled using Artificial Neural Microcircuits (ANMs), intended as off-the-shelf components, is articulated; the results of initial work to produce a catalogue of such Microcircuits though the use of Novelty Search is shown; followed by efforts to expand upon this initial work, including a discussion of challenges uncovered during these efforts and explorations of methods by which they might be overcome.",

keywords = "artificial neural networks, neural microcircuits, neuromorphic systems, fault tolerance, electronic systems",

author = "Andrew Walter and Shimeng Wu and Andy Tyrrell and Liam McDaid and Malachy McElholm and {Thandassery Sumithran}, Nidhin and Jim Harkin and Trefzer, {Martin Albrecht}",

year = "2024",

month = mar,

day = "24",

doi = "10.48550/arXiv.2403.16327",

language = "English",

publisher = "arXiv",

type = "WorkingPaper",

institution = "arXiv",

}

TY - UNPB

T1 - Artificial Neural Microcircuits as Building Blocks: Concept and Challenges

AU - Walter, Andrew

AU - Wu, Shimeng

AU - Tyrrell, Andy

AU - McDaid, Liam

AU - McElholm, Malachy

AU - Thandassery Sumithran, Nidhin

AU - Harkin, Jim

AU - Trefzer, Martin Albrecht

PY - 2024/3/24

Y1 - 2024/3/24

N2 - Artificial Neural Networks (ANNs) are one of the most widely employed forms of bio-inspired computation. However the current trend is for ANNs to be structurally homogeneous. Furthermore, this structural homogeneity requires the application of complex training and learning tools that produce application specific ANNs, susceptible to pitfalls such as overfitting. In this paper, an new approach is explored, inspired by the role played in biology by Neural Microcircuits, the so called ``fundamental processing elements'' of organic nervous systems. How large neural networks, particularly Spiking Neural Networks (SNNs) can be assembled using Artificial Neural Microcircuits (ANMs), intended as off-the-shelf components, is articulated; the results of initial work to produce a catalogue of such Microcircuits though the use of Novelty Search is shown; followed by efforts to expand upon this initial work, including a discussion of challenges uncovered during these efforts and explorations of methods by which they might be overcome.

AB - Artificial Neural Networks (ANNs) are one of the most widely employed forms of bio-inspired computation. However the current trend is for ANNs to be structurally homogeneous. Furthermore, this structural homogeneity requires the application of complex training and learning tools that produce application specific ANNs, susceptible to pitfalls such as overfitting. In this paper, an new approach is explored, inspired by the role played in biology by Neural Microcircuits, the so called ``fundamental processing elements'' of organic nervous systems. How large neural networks, particularly Spiking Neural Networks (SNNs) can be assembled using Artificial Neural Microcircuits (ANMs), intended as off-the-shelf components, is articulated; the results of initial work to produce a catalogue of such Microcircuits though the use of Novelty Search is shown; followed by efforts to expand upon this initial work, including a discussion of challenges uncovered during these efforts and explorations of methods by which they might be overcome.

KW - artificial neural networks

KW - neural microcircuits

KW - neuromorphic systems

KW - fault tolerance

KW - electronic systems

U2 - 10.48550/arXiv.2403.16327

DO - 10.48550/arXiv.2403.16327

M3 - Preprint

BT - Artificial Neural Microcircuits as Building Blocks: Concept and Challenges

PB - arXiv

ER -