Research map / Architectures and efficiency / Spiking networks
IM-Loss: Information Maximization Loss for Spiking Neural Networks
Yufei Guo, Yuanpei Chen, Liwen Zhang, Xiaode Liu, Yinglei Wang, Xuhui Huang, Zhe Ma
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 |
|---|---|---|---|
| Add IM-Loss to CIFARNet without normalization (ablation) | spiking CIFARNet · CIFAR10 | accuracy: 89.83% -> 90.90% | Improved |
| Add ESG to CIFARNet without normalization (ablation) | spiking CIFARNet · CIFAR10 | accuracy: 89.83% -> 91.17% | Improved |
| Combine IM-Loss and ESG on CIFARNet (ablation) | spiking CIFARNet · CIFAR10 | accuracy: 90.90% -> 91.75% | Improved |
| Add tdBN on top of IM-Loss and ESG for CIFARNet (ablation) | spiking CIFARNet · CIFAR10 | accuracy: 91.75% -> 92.20% | Improved |
| Add IM-Loss to ResNet-19 without normalization (ablation) | spiking ResNet-19 · CIFAR10 | accuracy: 91.23% -> 94.29% | Improved |
| Add ESG to ResNet-19 without normalization (ablation) | spiking ResNet-19 · CIFAR10 | accuracy: 91.23% -> 94.44% | Improved |
| Combine IM-Loss and ESG on ResNet-19 (ablation) | spiking ResNet-19 · CIFAR10 | accuracy: 94.29% -> 94.64% | Improved |
| Add tdBN on top of IM-Loss and ESG for ResNet-19 (ablation) | spiking ResNet-19 · CIFAR10 | accuracy: 94.64% -> 95.40% | Improved |
| Replace tdBN with IM-Loss on ResNet-19 (ablation) | spiking ResNet-19 · CIFAR10 | accuracy: 92.92% -> 94.29% | Improved |
| Use ESG (K from 100 to 101) instead of fixed k=100 or k=101 (ablation) | spiking CIFARNet and ResNet-19 · CIFAR10, timestep 4 | convergence and accuracy | Improved |
| Larger fixed surrogate k=101 instead of k=100 (ablation) | spiking CIFARNet and ResNet-19 · CIFAR10 | convergence | Hurt |
| Reduce timesteps from 6 to 2 with IM-Loss+ESG on ResNet-19 | IM-Loss+ESG ResNet-19 · CIFAR10 | accuracy: 95.49% -> 93.85% | Hurt |
Closest papers by meaning
- Take A Shortcut Back: Mitigating the Gradient Vanishing for Training Spiking Neural Networks (NeurIPS 2024 · similarity 0.89)
- Spiking Neural Networks Need High-Frequency Information (NeurIPS 2025 · similarity 0.89)
- EnOF-SNN: Training Accurate Spiking Neural Networks via Enhancing the Output Feature (NeurIPS 2024 · similarity 0.88)
- A Scalable, Causal, and Energy Efficient Framework for Neural Decoding with Spiking Neural Networks (NeurIPS 2025 · similarity 0.88)
- S$^2$NN: Sub-bit Spiking Neural Networks (NeurIPS 2025 · similarity 0.88)
- Rethinking the Membrane Dynamics and Optimization Objectives of Spiking Neural Networks (NeurIPS 2024 · similarity 0.88)
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.
