使用带自突触的单个神经元重建脉冲神经网络
Reconstructing Spiking Neural Networks Using a Single Neuron with Autapses
用带时延自突触的单个LIF神经元通过时序复用实现储备池、MLP与卷积类脉冲架构，为视觉任务提供一种以时间延迟换取神经元数量与状态显存大幅下降的紧凑计算单元思路。
原文依据 TDA-SNN greatly reduces neuron count and state memory while increasing per-neuron information capacity
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论脉冲神经网络（SNN）在神经形态计算中具有前景，但高性能模型仍依赖密集的多层架构，带来可观的通信与状态存储开销。受自突触启发，我们提出 TDA-SNN，一个利用带时延自突触的单个漏积分放电神经元以及基于原型学习的训练策略来重建 SNN 架构的框架。通过重组内部时序状态，TDA-SNN 可在统一框架内实现储备池、多层感知机与类卷积的脉冲架构。在序列、事件驱动与图像基准上的实验表明，其在储备池与 MLP 设定下具有有竞争力的性能，而卷积结果则揭示出明显的时空权衡。与标准 SNN 相比，TDA-SNN 大幅减少神经元数量与状态存储，同时提升单神经元的信息容量，代价是在极端单神经元设定下增加额外的时序延迟。这些发现凸显了时序复用单神经元模型作为类脑计算紧凑计算单元的潜力。
英文原文摘要 1126 字符
Spiking neural networks (SNNs) are promising for neuromorphic computing, but high-performing models still rely on dense multilayer architectures with substantial communication and state-storage costs. Inspired by autapses, we propose TDA-SNN, a framework that reconstructs SNN architectures using a single leaky integrate-and-fire neuron with time-delayed autapses and a prototype-learning-based training strategy. By reorganizing internal temporal states, TDA-SNN can realize reservoir, multilayer perceptron, and convolution-like spiking architectures within a unified framework. Experiments on sequential, event-based, and image benchmarks show competitive performance in reservoir and MLP settings, while convolutional results reveal a clear space--time trade-off. Compared with standard SNNs, TDA-SNN greatly reduces neuron count and state memory while increasing per-neuron information capacity, at the cost of additional temporal latency in extreme single-neuron settings. These findings highlight the potential of temporally multiplexed single-neuron models as compact computational units for brain-inspired computing.