Purin: A Biology-inspired Mechanism for Artificial Neural Networks

2026-09-25

AI Summary

This study addresses the limitations of traditional artificial neural networks, which overlook synaptic short-term efficacy dynamics and rely on discrete time steps to process temporal features. We propose Purin, a biologically inspired mechanism that introduces synaptic efficacy modulation into convolutional neural networks. By pioneering an inter-spike interval-based abstraction method combined with bounded factors and dual weight matrices, Purin effectively simulates biological synaptic plasticity. This mechanism enables both short- and long-term dynamic adjustments through temporal interval abstraction, facilitating backpropagation training without requiring discrete time steps. Experimental results demonstrate that integrating Purin significantly improves classification accuracy in models such as AlexNet. Ultimately, this work establishes a novel paradigm for constructing deep neural networks with enhanced biological plausibility.

Abstract

Artificial neural networks (ANNs) usually represent neural transmission with fixed trainable weights during a training batch, which omits short-term changes in synaptic efficacy. In addition, the discrete time-step simulation requires additional temporal processing that many conventional ANN architectures do not use. To overcome these challenges, we propose Purin, a biology-inspired and ANN-compatible mechanism, that introduces synaptic efficacy modulation into conventional convolutional neural networks. Purin uses a time-interval-based abstraction for neural activities, which allows Purin to introduce short- and long-term synaptic efficacy changes without using discrete time-steps. Purin introduces a bounded factor to represent temporary synaptic efficacy changes, together with two weight matrices that represent input-side and output-side efficacy. The weight matrices are updated by backpropagation and interpreted as the long-term synaptic efficacy changes. Experimental results show that after removing the confounding factors in the AlexNet, VGG11, and GoogLeNet architectures, Purin improves the classification accuracies in all three models across the evaluated datasets.

Problem

• Artificial Neural Networks
• Synaptic Efficacy
• Temporal Processing
• Discrete Time-steps

Innovation

• Biology-inspired mechanism
• Synaptic efficacy modulation
• Time-interval-based abstraction
• Artificial neural networks
• Convolutional neural networks