Multi-plasticity synergy with adaptive mechanism assignment for training spiking neural networks
TL;DR
This work proposes a biology-motivated computational framework that incorporates multiple synergistic plasticity mechanisms for more effective SNN training and provides a general and extensible foundation for developing more powerful SNNs guided by multi-strategy brain-inspired learning.
Abstract
Spiking Neural Networks (SNNs) are promising brain-inspired models known for low power consumption and superior potential for temporal processing, but identifying suitable learning mechanisms remains a challenge. Despite the presence of multiple coexisting learning strategies in the brain, current SNN training methods typically rely on a single form of synaptic plasticity, which limits their adaptability and representational capability. In this paper, we propose a biology-motivated computational framework that incorporates multiple synergistic plasticity mechanisms for more effective SNN training. The framework is inspired by the coexistence of heterogeneous regulatory and plasticity-related processes in biological neural systems, rather than directly corresponding to specific biological learning rules. Our method enables diverse learning algorithms to cooperatively modulate the accumulation of information, while allowing each mechanism to preserve its own relatively independent update dynamics. We evaluated our approach on diverse datasets to demonstrate that our framework significantly improves performance and robustness compared to conventional learning mechanism models. This work provides a general and extensible foundation for developing more powerful SNNs guided by multi-strategy brain-inspired learning.