Research map / Architectures and efficiency / Spiking networks
RMP-SNN: Residual Membrane Potential Neuron for Enabling Deeper High-Accuracy and Low-Latency Spiking Neural Network
Check 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 |
|---|---|---|---|
| Replace hard reset with residual-potential soft reset | ResNet-20 SNN · CIFAR-10 | top-1 accuracy: 87.46% -> 91.36% | Helped |
| Replace hard reset with residual-potential soft reset | VGG-16 SNN · CIFAR-10 | top-1 accuracy: 91.55% -> 93.63% | Helped |
| Replace hard reset with residual-potential soft reset | ResNet-20 SNN · CIFAR-100 | top-1 accuracy: 64.09% -> 67.82% | Helped |
| Replace hard reset with residual-potential soft reset | VGG-16 SNN · CIFAR-100 | top-1 accuracy: 70.77% -> 70.93% | Helped |
| Replace hard reset with residual-potential soft reset | ResNet-34 SNN · ImageNet | top-1 accuracy: 65.47% -> 69.89% | Helped |
| Replace hard reset with residual-potential soft reset | VGG-16 SNN · ImageNet | top-1 accuracy: 69.96% -> 73.09% | Helped |
| Scale layer thresholds down from the maximum-input initialization (ablation) | ResNet-20 RMP-SNN · CIFAR-10 | top-1 accuracy: 91.36% -> same accuracy | No clear effect |
| Reduce thresholds further to 0.2× (ablation) | ResNet-20 RMP-SNN · CIFAR-10 | top-1 accuracy | Hurt |
| Use reduced thresholds for faster inference (ablation) | VGG-16 RMP-SNN · CIFAR-10 | inference time-steps to exceed 90% accuracy: 512 -> 64 time-steps | Helped |
| Use reduced thresholds for faster inference (ablation) | VGG-16 RMP-SNN · CIFAR-100 | inference time-steps to reach 68.34% accuracy: about 512 -> 256 time-steps | Helped |
| Use reduced thresholds for faster inference (ablation) | VGG-16 RMP-SNN · ImageNet | inference time-steps to reach 68.93% accuracy: over 2300 -> 512 time-steps | Helped |
| Use reduced thresholds for faster inference (ablation) | ResNet-20 RMP-SNN · CIFAR-10 | inference time-steps to exceed 85% accuracy: 512 -> 64 time-steps | Helped |
Closest papers by meaning
- Efficient ANN-SNN Conversion with Error Compensation Learning (ICML 2025 · similarity 0.91)
- Rethinking the Membrane Dynamics and Optimization Objectives of Spiking Neural Networks (NeurIPS 2024 · similarity 0.91)
- Adaptive Calibration: A Unified Conversion Framework of Spiking Neural Networks (AAAI 2025 · similarity 0.90)
- Inference-Scale Complexity in ANN-SNN Conversion for High-Performance and Low-Power Applications (CVPR 2025 · similarity 0.90)
- EnOF-SNN: Training Accurate Spiking Neural Networks via Enhancing the Output Feature (NeurIPS 2024 · similarity 0.90)
- Faster and Stronger: When ANN-SNN Conversion Meets Parallel Spiking Calculation (ICML 2025 · 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.