Papers › Differentiable Spike: Rethinking Gradient-Descent for Training Spiking Neural Networks

Differentiable Spike: Rethinking Gradient-Descent for Training Spiking Neural Networks

1 Dec 2021NeurIPS 2021 12archive 2025-07-28

Yuhang Li, Yufei Guo, Shanghang Zhang, Shikuang Deng, Yongqing Hai, Shi Gu

Spiking Neural Networks (SNNs) have emerged as a biology-inspired method mimicking the spiking nature of brain neurons. This bio-mimicry derives SNNs' energy efficiency of inference on neuromorphic hardware. However, it also causes an intrinsic disadvantage in training high-performing SNNs from scratch since the discrete spike prohibits the gradient calculation. To overcome this issue, the surrogate gradient (SG) approach has been proposed as a continuous relaxation. Yet the heuristic choice of SG leaves it vacant how the SG benefits the SNN training. In this work, we first theoretically study the gradient descent problem in SNN training and introduce finite difference gradient to quantitatively analyze the training behavior of SNN. Based on the introduced finite difference gradient, we propose a new family of Differentiable Spike (Dspike) functions that can adaptively evolve during training to find the optimal shape and smoothness for gradient estimation. Extensive experiments over several popular network structures show that training SNN with Dspike consistently outperforms the state-of-the-art training methods. For example, on the CIFAR10-DVS classification task, we can train a spiking ResNet-18 and achieve 75.4% top-1 accuracy with 10 time steps.

PaperPDFConference PDF

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Event data classificationImage Classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Event data classification CIFAR10-DVS Dspike (ResNet-18) Accuracy 75.4 #4 of 7 Archive leaderboard report
Image Classification CIFAR-100 Dspike (ResNet-18) Percentage correct 74.24 #155 of 211 Archive leaderboard report
Image Classification ImageNet Dspike (VGG-16) Top 1 Accuracy 71.24 #1012 of 1060 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections