Online Stabilization of Spiking Neural Networks

Authors: Yaoyu Zhu, Jianhao Ding, Tiejun Huang, Xiaodong Xie, Zhaofei Yu

ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments conducted on various datasets demonstrate the proposed method s superior performance among SNN online training algorithms. Our code is available at https://github.com/zhuyaoyu/SNN-onlinenormalization.
Researcher Affiliation Academia Yaoyu Zhu1, Jianhao Ding1, Tiejun Huang1,2, Xiaodong Xie1 & Zhaofei Yu1,2 1 School of Computer Science, Peking University 2 Institute for Artificial Intelligence, Peking University
Pseudocode Yes The overall algorithm description is provided in Appendix B.
Open Source Code Yes Our code is available at https://github.com/zhuyaoyu/SNN-onlinenormalization.
Open Datasets Yes We conduct experiments on CIFAR10, CIFAR100 (Krizhevsky et al., 2009), DVS-Gesture (Amir et al., 2017), CIFAR10-DVS (Li et al., 2017), and Imagenet (Deng et al., 2009) datasets to evaluate the performance of our method.
Dataset Splits No The paper describes data augmentation for training and mentions the test sets for evaluation, but does not explicitly provide details about specific validation dataset splits (percentages, counts, or explicit standard splits for a validation set).
Hardware Specification Yes All experiments are run on Nvidia RTX 4090 GPUs with Pytorch 2.0.
Software Dependencies Yes All experiments are run on Nvidia RTX 4090 GPUs with Pytorch 2.0.
Experiment Setup Yes Other hyperparameters we use are provided in Table 3, including total training epochs, batch size, learning rate, weight decay, ϵ (weight of MSE loss in Eq. 5), and dropout rate.