Publication

Neuromorphic learning, working memory, and metaplasticity in nanowire networks

MDR Open Deposited
Analytics

Downloadable Content

Download PDF
Download This Work (Zip)
Citations

Nanowire networks (NWNs) mimic the brain’s neurosynaptic connectivity and emergent dynamics. Consequently, NWNs may also emulate the synaptic processes that enable higher-order cognitive functions such as learning and memory. A quintessential cognitive task used to measure human working memory is the n-back task. In this study, task variations inspired by the n-back task are implemented in a NWN device, and external feedback is applied to emulate brain-like supervised and reinforcement learning. NWNs are found to retain information in working memory to at least n = 7 steps back, remarkably similar to the originally proposed “seven plus or minus two” rule for human subjects. Simulations elucidate how synapse-like NWN junction plasticity depends on previous synaptic modifications, analogous to “synaptic metaplasticity” in the brain, and how memory is consolidated via strengthening and pruning of synaptic conductance pathways.

First published at
Creator
  • Name
    Alon Loeffler
    Organization
    The University of Sydney
    Sub organization
    School of Physics
    Role
    author
    Name
    Adrian Diaz-Alvarez
    Organization
    NIMS
    Sub organization
    International Center for Young Scientists (ICYS)
    Role
    author
    Name
    Ruomin Zhu
    Organization
    The University of Sydney
    Sub organization
    School of Physics
    Role
    author
    Name
    Natesh Ganesh
    Organization
    National Institute of Standards and Technology (NIMS)
    Role
    author
    Name
    James M. Shine
    Organization
    The University of Sydney
    Sub organization
    School of Physics
    Role
    author
    Name
    Tomonobu Nakayama
    ORCID
    https://orcid.org/0000-0001-9696-475X
    Organization
    NIMS
    Sub organization
    International Center for Materials Nanoarchitectonics (MANA)
    Role
    author
    Name
    Zdenka Kuncic
    Organization
    The University of Sydney
    Sub organization
    School of Physics
    Role
    author
Keyword
Resource type
Publisher
Date published
  • 21/04/2023
Rights statement
Journal
Language
Last modified
  • 28/04/2023

Items

Thumbnail Title Date Uploaded Size Visibility Actions
2023_Alon_Diaz-Alvarez_Sci.Adv._sciadv.adg3289.pdf 1.83 MB MDR Open
PhysicalReinforcementLearning.rar 28.9 MB MDR Open
sciadv.adg3289_movies_s1_to_s4.zip 22.6 MB MDR Open
sciadv.adg3289_sm.pdf 6.5 MB MDR Open
sciadv.adg3289_source_data_file_s1.zip 31.8 KB MDR Open