Online Pseudo-Zeroth-Order Training of Neuromorphic Spiking Neural Networks

TL;DR

This paper introduces an online pseudo-zeroth-order training method for spiking neural networks that simplifies training by requiring only a single forward pass with noise, making it more hardware-friendly and biologically plausible.

Contribution

The proposed OPZO method enables effective online training of SNNs without spatial backpropagation, addressing variance issues and improving hardware compatibility.

Findings

OPZO achieves comparable accuracy to spatial BP on various datasets.

The method demonstrates low training costs and suitability for on-chip SNN training.

OPZO effectively handles variance issues in zeroth-order optimization.

Abstract

Brain-inspired neuromorphic computing with spiking neural networks (SNNs) is a promising energy-efficient computational approach. However, successfully training SNNs in a more biologically plausible and neuromorphic-hardware-friendly way is still challenging. Most recent methods leverage spatial and temporal backpropagation (BP), not adhering to neuromorphic properties. Despite the efforts of some online training methods, tackling spatial credit assignments by alternatives with comparable performance as spatial BP remains a significant problem. In this work, we propose a novel method, online pseudo-zeroth-order (OPZO) training. Our method only requires a single forward propagation with noise injection and direct top-down signals for spatial credit assignment, avoiding spatial BP's problem of symmetric weights and separate phases for layer-by-layer forward-backward propagation. OPZO solves the large variance problem of zeroth-order methods by the pseudo-zeroth-order formulation and momentum feedback connections, while having more guarantees than random feedback. Combining online training, OPZO can pave paths to on-chip SNN training. Experiments on neuromorphic and static datasets with fully connected and convolutional networks demonstrate the effectiveness of OPZO with similar performance compared with spatial BP, as well as estimated low training costs.