Research map / Architectures and efficiency / Spiking networks
UniSparse: Combining Weight Pruning and Spike Sparsification in Spiking Neural Networks
Xinyu Shi, Tong Bu, Zhaofei Yu
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 |
|---|---|---|---|
| use UniSparse instead of UPF | UPF · CIFAR-10 CifarNet lowest-SOPs point | accuracy: 90.65 -> 90.76 | Improved |
| use UniSparse ResNet-18 instead of UPF | UPF · CIFAR-10 lowest-SOPs point | accuracy: 90.65 -> 92.38 | Improved |
| use UniSparse ResNet-19 instead of UPF | UPF · CIFAR-10 lowest-SOPs point | accuracy: 90.65 -> 91.56 | Improved |
| use UniSparse instead of UPF | UPF · DVS-CIFAR10 | accuracy: 78.3 -> 79.12 | Improved |
| use UniSparse instead of UPF | UPF · ImageNet ResNet-18 | accuracy: 58.99 -> 59.66 | Improved |
| add weight pruning to spike sparsification only (ablation) | UniSparse (spike sparsification only) · DVS-CIFAR10 | average SOPs | Improved |
| add spike sparsification to weight pruning only (ablation) | UniSparse (weight pruning only) · DVS-CIFAR10 | accuracy at matched SOPs | Improved |
| apply UniSparse to Spikformer-8-384 spiking transformer | Spikformer-8-384 · ImageNet | accuracy: 72.21 -> 70.68 -> 68.29 | No clear effect |
Closest papers by meaning
- Improving the Sparse Structure Learning of Spiking Neural Networks from the View of Compression Efficiency (ICLR 2025 · similarity 0.89)
- Adaptive Calibration: A Unified Conversion Framework of Spiking Neural Networks (AAAI 2025 · similarity 0.89)
- QP-SNN: Quantized and Pruned Spiking Neural Networks (ICLR 2025 · similarity 0.89)
- Towards efficient deep spiking neural networks construction with spiking activity based pruning (ICML 2024 · similarity 0.89)
- S$^2$NN: Sub-bit Spiking Neural Networks (NeurIPS 2025 · similarity 0.89)
- Sparse Spiking Neural Network: Exploiting Heterogeneity in Timescales for Pruning Recurrent SNN (ICLR 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.
