Characterizing Learning in Spiking Neural Networks with Astrocyte-Like Units

2025-03-09

AI Summary

The regulatory role of astrocyte-like units in spiking neural networks (SNNs) and the impact of astrocyte–neuron ratio on learning performance remain poorly understood. Method: We propose a hybrid SNN architecture inspired by liquid state machines (LSMs), integrating biologically plausible astrocytic dynamics for chaotic time-series prediction. Contribution/Results: We demonstrate for the first time that astrocyte-like units significantly modulate synaptic plasticity and temporal information integration. Empirical results show that a 2:1 astrocyte-to-neuron ratio yields peak learning speed, along with superior prediction accuracy and robustness compared to purely neuronal or purely astrocytic networks. Notably, this optimal ratio closely matches the empirically observed astrocyte–neuron ratio in mammalian brains, providing strong evidence for astrocytes’ critical role in long-term dynamic representation and cooperative learning. Our work establishes a novel, biologically interpretable neuromorphic computing paradigm grounded in glial–neuronal co-processing principles.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Neuro-Symbolic Learning
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Social networks and social learning

Abstract

Traditional artificial neural networks take inspiration from biological networks, using layers of neuron-like nodes to pass information for processing. More realistic models include spiking in the neural network, capturing the electrical characteristics more closely. However, a large proportion of brain cells are of the glial cell type, in particular astrocytes which have been suggested to play a role in performing computations. Here, we introduce a modified spiking neural network model with added astrocyte-like units in a neural network and asses their impact on learning. We implement the network as a liquid state machine and task the network with performing a chaotic time-series prediction task. We varied the number and ratio of neuron-like and astrocyte-like units in the network to examine the latter units effect on learning. We show that the combination of neurons and astrocytes together, as opposed to neural- and astrocyte-only networks, are critical for driving learning. Interestingly, we found that the highest learning rate was achieved when the ratio between astrocyte-like and neuron-like units was roughly 2 to 1, mirroring some estimates of the ratio of biological astrocytes to neurons. Our results demonstrate that incorporating astrocyte-like units which represent information across longer timescales can alter the learning rates of neural networks, and the proportion of astrocytes to neurons should be tuned appropriately to a given task.

Problem

• Impact of astrocyte-like units on learning in spiking neural networks.
• Optimal ratio of astrocyte-like to neuron-like units for learning.
• Role of astrocytes in enhancing neural network learning rates.

Innovation

• Incorporated astrocyte-like units in spiking neural networks
• Used liquid state machine for chaotic time-series prediction
• Optimized astrocyte-to-neuron ratio for enhanced learning rates