Research map / Architectures and efficiency / Spiking networks
Hybrid Macro/Micro Level Backpropagation for Training Deep Spiking Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use HM2-BP instead of STBP for the fully connected SNN | Spiking MLP on MNIST · MNIST | accuracy: 98.89% vs 98.84 ± 0.02% | No clear effect |
| Use HM2-BP instead of STBP | Spiking MLP · N-MNIST | accuracy: 98.78% vs 98.84 ± 0.02% | Helped |
| Use HM2-BP instead of STBP | Spiking CNN · MNIST | accuracy: 99.42% vs 99.42 ± 0.11% | No clear effect |
| Use HM2-BP instead of eRBP | Spiking MLP, 200-200 · EMNIST | accuracy: 78.17% -> 84.31 ± 0.10% | Helped |
| Use HM2-BP instead of ANN training | 200-200 network · EMNIST | accuracy: 81.77% -> 84.31 ± 0.10% | Helped |
| Remove data augmentation (ablation) | Spiking CNN (HM2-BP) · MNIST | accuracy: 99.42 ± 0.11% -> 99.32 ± 0.05% | Hurt |
| Use an 800-800 architecture rather than an 800-unit single hidden layer | HM2-BP · TI46 Speech, 16 speakers | accuracy: 89.36± 0.30% -> 90.50± 0.45% | Helped |
Closest papers by meaning
- Spike-Train Level Backpropagation for Training Deep Recurrent Spiking Neural Networks (NeurIPS 2019 · similarity 0.92)
- Enabling Deep Spiking Neural Networks with Hybrid Conversion and Spike Timing Dependent Backpropagation (ICLR 2020 · similarity 0.92)
- Temporal Spike Sequence Learning via Backpropagation for Deep Spiking Neural Networks (NeurIPS 2020 · similarity 0.91)
- Advancing Training Efficiency of Deep Spiking Neural Networks through Rate-based Backpropagation (NeurIPS 2024 · similarity 0.90)
- Enhanced Self-Distillation Framework for Efficient Spiking Neural Network Training (NeurIPS 2025 · similarity 0.89)
- Take A Shortcut Back: Mitigating the Gradient Vanishing for Training Spiking Neural Networks (NeurIPS 2024 · similarity 0.89)
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.