Toward Efficient Spiking Transformers: Synapse Pruning Meets Synergistic Learning-Based Compensation
By: Hongze Sun , Wuque Cai , Duo Chen and more
Potential Business Impact:
Makes AI smarter, smaller, and faster.
As a foundational architecture of artificial intelligence models, Transformer has been recently adapted to spiking neural networks with promising performance across various tasks. However, existing spiking Transformer (ST)-based models require a substantial number of parameters and incur high computational costs, thus limiting their deployment in resource-constrained environments. To address these challenges, we propose combining synapse pruning with a synergistic learning-based compensation strategy to derive lightweight ST-based models. Specifically, two types of tailored pruning strategies are introduced to reduce redundancy in the weight matrices of ST blocks: an unstructured $\mathrm{L_{1}P}$ method to induce sparse representations, and a structured DSP method to induce low-rank representations. In addition, we propose an enhanced spiking neuron model, termed the synergistic leaky integrate-and-fire (sLIF) neuron, to effectively compensate for model pruning through synergistic learning between synaptic and intrinsic plasticity mechanisms. Extensive experiments on benchmark datasets demonstrate that the proposed methods significantly reduce model size and computational overhead while maintaining competitive performance. These results validate the effectiveness of the proposed pruning and compensation strategies in constructing efficient and high-performing ST-based models.
Similar Papers
A flexible framework for structural plasticity in GPU-accelerated sparse spiking neural networks
Neural and Evolutionary Computing
Makes computer brains learn faster and use less power.
Multi-Plasticity Synergy with Adaptive Mechanism Assignment for Training Spiking Neural Networks
Neural and Evolutionary Computing
Teaches computer brains to learn better, faster.
Spatio-Temporal Pruning for Compressed Spiking Large Language Models
Neural and Evolutionary Computing
Makes smart computer brains use less power.