Cargando…
A bio-inspired bistable recurrent cell allows for long-lasting memory
Recurrent neural networks (RNNs) provide state-of-the-art performances in a wide variety of tasks that require memory. These performances can often be achieved thanks to gated recurrent cells such as gated recurrent units (GRU) and long short-term memory (LSTM). Standard gated cells share a layer in...
Autores principales: | , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Public Library of Science
2021
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8186810/ https://www.ncbi.nlm.nih.gov/pubmed/34101750 http://dx.doi.org/10.1371/journal.pone.0252676 |
_version_ | 1783705020994158592 |
---|---|
author | Vecoven, Nicolas Ernst, Damien Drion, Guillaume |
author_facet | Vecoven, Nicolas Ernst, Damien Drion, Guillaume |
author_sort | Vecoven, Nicolas |
collection | PubMed |
description | Recurrent neural networks (RNNs) provide state-of-the-art performances in a wide variety of tasks that require memory. These performances can often be achieved thanks to gated recurrent cells such as gated recurrent units (GRU) and long short-term memory (LSTM). Standard gated cells share a layer internal state to store information at the network level, and long term memory is shaped by network-wide recurrent connection weights. Biological neurons on the other hand are capable of holding information at the cellular level for an arbitrary long amount of time through a process called bistability. Through bistability, cells can stabilize to different stable states depending on their own past state and inputs, which permits the durable storing of past information in neuron state. In this work, we take inspiration from biological neuron bistability to embed RNNs with long-lasting memory at the cellular level. This leads to the introduction of a new bistable biologically-inspired recurrent cell that is shown to strongly improves RNN performance on time-series which require very long memory, despite using only cellular connections (all recurrent connections are from neurons to themselves, i.e. a neuron state is not influenced by the state of other neurons). Furthermore, equipping this cell with recurrent neuromodulation permits to link them to standard GRU cells, taking a step towards the biological plausibility of GRU. With this link, this work paves the way for studying more complex and biologically plausible neuromodulation schemes as gating mechanisms in RNNs. |
format | Online Article Text |
id | pubmed-8186810 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2021 |
publisher | Public Library of Science |
record_format | MEDLINE/PubMed |
spelling | pubmed-81868102021-06-16 A bio-inspired bistable recurrent cell allows for long-lasting memory Vecoven, Nicolas Ernst, Damien Drion, Guillaume PLoS One Research Article Recurrent neural networks (RNNs) provide state-of-the-art performances in a wide variety of tasks that require memory. These performances can often be achieved thanks to gated recurrent cells such as gated recurrent units (GRU) and long short-term memory (LSTM). Standard gated cells share a layer internal state to store information at the network level, and long term memory is shaped by network-wide recurrent connection weights. Biological neurons on the other hand are capable of holding information at the cellular level for an arbitrary long amount of time through a process called bistability. Through bistability, cells can stabilize to different stable states depending on their own past state and inputs, which permits the durable storing of past information in neuron state. In this work, we take inspiration from biological neuron bistability to embed RNNs with long-lasting memory at the cellular level. This leads to the introduction of a new bistable biologically-inspired recurrent cell that is shown to strongly improves RNN performance on time-series which require very long memory, despite using only cellular connections (all recurrent connections are from neurons to themselves, i.e. a neuron state is not influenced by the state of other neurons). Furthermore, equipping this cell with recurrent neuromodulation permits to link them to standard GRU cells, taking a step towards the biological plausibility of GRU. With this link, this work paves the way for studying more complex and biologically plausible neuromodulation schemes as gating mechanisms in RNNs. Public Library of Science 2021-06-08 /pmc/articles/PMC8186810/ /pubmed/34101750 http://dx.doi.org/10.1371/journal.pone.0252676 Text en © 2021 Vecoven et al https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. |
spellingShingle | Research Article Vecoven, Nicolas Ernst, Damien Drion, Guillaume A bio-inspired bistable recurrent cell allows for long-lasting memory |
title | A bio-inspired bistable recurrent cell allows for long-lasting memory |
title_full | A bio-inspired bistable recurrent cell allows for long-lasting memory |
title_fullStr | A bio-inspired bistable recurrent cell allows for long-lasting memory |
title_full_unstemmed | A bio-inspired bistable recurrent cell allows for long-lasting memory |
title_short | A bio-inspired bistable recurrent cell allows for long-lasting memory |
title_sort | bio-inspired bistable recurrent cell allows for long-lasting memory |
topic | Research Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8186810/ https://www.ncbi.nlm.nih.gov/pubmed/34101750 http://dx.doi.org/10.1371/journal.pone.0252676 |
work_keys_str_mv | AT vecovennicolas abioinspiredbistablerecurrentcellallowsforlonglastingmemory AT ernstdamien abioinspiredbistablerecurrentcellallowsforlonglastingmemory AT drionguillaume abioinspiredbistablerecurrentcellallowsforlonglastingmemory AT vecovennicolas bioinspiredbistablerecurrentcellallowsforlonglastingmemory AT ernstdamien bioinspiredbistablerecurrentcellallowsforlonglastingmemory AT drionguillaume bioinspiredbistablerecurrentcellallowsforlonglastingmemory |