INTRO: The brain contains billions of astrocytes, and scientists have long known they play a part in cleaning up molecules within brain synapses, the junctions where neurons come together.
Little-known cells might be key to human brain’s massive memory
A
new model suggests that astrocytes might be used in computation,
coordinating with neurons and connecting synapses in networks
The new model suggests the astrocytes could also be used for computation, coordinating with neurons and connecting synapses in networks.
These complex connections might allow the brain to encode memories in
dense networks that expand the brain’s capacity beyond its neurons
alone, the researchers write.
“This makes neuron-astrocyte networks an
exciting candidate for biological ‘hardware’ implementing Dense Associative Memory,” they add.
- The model runs counter to the prevailing theory that memory storage occurs in synapses, and implies that the brain is capable of storing even more memories than once thought possible.
- The researchers also describe how the theory could be validated in the lab.
Little-known cells might be key to human brain’s massive memory
Story by
A new model of memory — and a little-heralded type of brain cell — might explain why the human brain has such a huge storage capacity, researchers reported in the journal PNAS in May. . .
We hope that one of the consequences of this work could be that experimentalists would consider this idea seriously and perform some experiments testing this hypothesis,” said Dmitry Krotov, a research staff member at the MIT-IBM Watson AI Lab and IBM Research and the paper’s senior author, in a news release.
The model could also be used as a “fresh source of inspiration” for future artificial intelligence technology, the researchers conclude.
Neuron–astrocyte associative memory
Edited
by Wesley P. Clawson, Allen Discovery Center at Tufts University,
Medford, MA; received September 11, 2024; accepted April 7, 2025 by
Editorial Board Member Michael S. Gazzaniga
Data, Materials, and Software Availability
Code data have been deposited in GitHub (https://github.com/kozleo/naam) (58). All other data are included in the manuscript and/or SI Appendix.
5. Discussion
We
have introduced a biologically inspired model that describes the
interactions between neurons, synapses, and astrocytes.
In our model,
astrocytes are able to adaptively control synaptic weights in an online
fashion.
Theoretical analysis has demonstrated that this model can
exhibit associative memory
behavior and is closely related to the Dense
Associative Memory family of models with supralinear memory capacity, as
well as to transformers.
We have shown that, through the choice of the
connectivity tensor,
our neuron–astrocyte model can be smoothly dialed between operating as a
transformer and operating as a Dense Associative Memory network. This
opens up the possibility for exploring novel architectures “in-between”
transformers and Dense Associative Memories. Furthermore, we have
presented a simple algorithm for memory storage and have provided
numerical evidence of our models’ effectiveness, such as successfully
storing and retrieving CIFAR10 and ImageNet images.
In
broader terms, this work proposes that memories can, at least in part,
be stored within the molecular machinery of astrocytes. This contrasts
with the prevailing neuroscience viewpoint that memories are stored in
the synaptic weights between neurons. To experimentally validate this
claim, one would need to selectively interfere with the ability of
to diffuse intracellularly through astrocytes. Our model predicts that
hindering this diffusion would significantly impair memory recall. Our
model is flexible enough to accommodate many different types of
process-to-process coupling patterns, which could presumably be fit to
match physiological data. For example, it is possible to enforce
“nearest-neighbor” coupling between astrocyte processes (which can be
achieved by e.g., imposing a block-diagonal structure on the tensor such that if processes and are not spatially close to each other), while still guaranteeing convergence of our model to a fixed point.
The
feedback loop between neurons and astrocytes is well established
experimentally, yet a simple computational account of its function
remains largely absent. Previous models have primarily focused on the
biophysical details of neuron–astrocyte interactions, aiming to capture
the emergence of complex spatiotemporal calcium dynamics (52, 53).
Building on these efforts, our work highlights the role of astrocytic
modulation in synaptic plasticity. However, rather than modeling
detailed biophysical mechanisms, we take a higher-level approach, using a
“firing rate” model to abstract away these complexities and uncover the
core computational principles governing neuron–astrocyte interactions.
One
particularly intriguing feature of astrocytes is their abundance in the
brain. Indeed, they are present in virtually every major brain
structure. Of particular relevance to the modeling approach introduced
here are associative brain areas, such as the neocortex and hippocampus,
which are believed to play a crucial role in memory storage and
retrieval. Notably, human neocortical astrocytes are significantly
larger and more active than their rodent counterparts, suggesting an
enhanced computational function (54). As in prior modeling work using recurrent networks (55),
our neuron–astrocyte model can be tuned (i.e., trained) to perform
functions associated with specific tasks relevant to particular brain
regions. For example, astrocytes in the visual cortex may be specialized
for storing and retrieving visual information, whereas those in
language-related areas may be more attuned to processing auditory
information.
While our focus has been on a
minicircuit consisting of a single astrocyte interacting with multiple
nearby synapses, astrocytes also extensively communicate with each other
through chemical gap junctions. Exploring the implications of this
intercellular coupling will be the subject of future research.
Key
ideas in machine learning and AI drew initial inspiration from
neuroscience, including neural networks, convolutional nets, threshold
linear (ReLu) units, and dropout.
Yet it is debatable whether
neuroscience research from the last fifty years has significantly
influenced or informed machine learning. Astrocytes, along with other
biological structures such as dendrites (56) and neuromodulators (57) may offer a fresh source of inspiration for building state-of-the-art AI systems.
SCREENGRAB
No comments:
Post a Comment