Research map / Architectures and efficiency / Spiking networks
Zeroth-Order Forward-Only SNN Training Inspiring Neuromorphic On-Chip Learning
Mingyue Qin, Shuyu Yin, Qinghai Guo, Peilin Liu, Xiaolin Huang, Fei Wen
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| apply zeroth-order training to an SNN instead of an ANN (Heaviside spiking activation replacing ReLU) | DeepZero (CGE) on ResNet20 · CIFAR10 | ZO training success | Hurt |
| train in a low-dimensional PCA subspace of the BP trajectory instead of full space | BP on Spike-ResNet50 · ImageNet1K | training speed: fewer than 10 epochs of subspace training match 60 epochs of full-space BP | Improved |
| use SZO-CGE instead of SZO-RGE in the low-dimensional subspace (ablation) | SZO · training from scratch | accuracy | No clear effect |
| increase number of queries q in SZO-RGE (q=1 to q=20) (ablation) | SZO-RGE · CIFAR10 Spike-ResNet20 | accuracy (%): 81.29 -> 83.71 | Improved |
| use SZO-CGE instead of full-space BP in continual training | Spike-ResNet18 · CIFAR100 continual training | max accuracy (%): 74.42 -> 74.59 | Improved |
| use subspace-based SZO instead of Tent with full-space BP for test-time adaptation (Continual) | Spike-ResNet50 · ImageNetC severity 5 | accuracy (%): 3.89 -> 33.37 | Improved |
| vary perturbation scale mu from 0.6 to 2.2 (ablation) | SZO-CGE · CIFAR10 / CIFAR100 from scratch | accuracy (%): 85.91 -> 85.92 (CIFAR10, mu 0.6 -> 2.2) | No clear effect |
| increase subspace dimension (ablation) | SZO-CGE · CIFAR100 Spike-ResNet18 from scratch | accuracy (%): 70.97 -> 73.87 (dim 20 -> 60, first window) | Improved |
| replace subspace ZO with feedback-based OPZO or DFA on deep ResNets | OPZO / DFA baselines · CIFAR100 Spike-ResNet18 | accuracy (%): 73.35 -> 36.90 (OPZO) | Hurt |
| use SZO instead of BP for memory | Spike-ResNet50 · ImageNet1K GPU memory | GPU memory | Improved |
Closest papers by meaning
- Spike-driven Transformer V2: Meta Spiking Neural Network Architecture Inspiring the Design of Next-generation Neuromorphic Chips (ICLR 2024 · similarity 0.87)
- Enhanced Self-Distillation Framework for Efficient Spiking Neural Network Training (NeurIPS 2025 · similarity 0.87)
- Temporal Flexibility in Spiking Neural Networks: Towards Generalization Across Time Steps and Deployment Friendliness (ICLR 2025 · similarity 0.86)
- NDOT: Neuronal Dynamics-based Online Training for Spiking Neural Networks (ICML 2024 · similarity 0.86)
- Rethinking the Membrane Dynamics and Optimization Objectives of Spiking Neural Networks (NeurIPS 2024 · similarity 0.86)
- Multiplication-Free Parallelizable Spiking Neurons with Efficient Spatio-Temporal Dynamics (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
