Delving Deeper Into Astromorphic Transformers

TL;DR

This paper explores bio-inspired neuron-astrocyte interactions to develop Astromorphic Transformers, improving accuracy, learning speed, and natural language generation capabilities across various machine learning tasks.

Contribution

It introduces a bioplausible model of neuron-astrocyte interactions to mimic self-attention in Transformers, integrating bio-realistic effects into machine learning applications.

Findings

Improved accuracy and learning speed on sentiment and image classification tasks.

Enhanced perplexity and generalization in natural language generation.

Demonstrated stability across diverse ML tasks.

Abstract

Preliminary attempts at incorporating the critical role of astrocytes - cells that constitute more than 50\% of human brain cells - in brain-inspired neuromorphic computing remain in infancy. This paper seeks to delve deeper into various key aspects of neuron-synapse-astrocyte interactions to mimic self-attention mechanisms in Transformers. The cross-layer perspective explored in this work involves bioplausible modeling of Hebbian and presynaptic plasticities in neuron-astrocyte networks, incorporating effects of non-linearities and feedback along with algorithmic formulations to map the neuron-astrocyte computations to self-attention mechanism and evaluating the impact of incorporating bio-realistic effects from the machine learning application side. Our analysis on sentiment and image classification tasks (IMDB and CIFAR10 datasets) highlights the advantages of Astromorphic Transformers, offering improved accuracy and learning speed. Furthermore, the model demonstrates strong natural language generation capabilities on the WikiText-2 dataset, achieving better perplexity compared to conventional models, thus showcasing enhanced generalization and stability across diverse machine learning tasks.