Research map / Architectures and efficiency / Spiking networks
Robust Spiking Neural Networks by Temporal Mutual Information
Mengting Xu, Shi Gu, Peng Lin, De Ma, Huajin Tang, Qian Zheng, Gang Pan
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 TMI regularizer to STBP training | STBP VGGSNN · CIFAR-100, FGSM | accuracy (%): 13.51 -> 15.20 | Improved |
| add TMI regularizer to TET training | TET VGGSNN · CIFAR-100, PGD | accuracy (%): 2.25 -> 4.63 | Improved |
| add TMI regularizer to SNN-RAT adversarial training | SNN-RAT VGG11 · CIFAR-100, FGSM | accuracy (%): 25.86 -> 27.32 | Improved |
| add TMI regularizer | STBP VGGSNN · CIFAR-100, natural accuracy | accuracy (%): 68.11 -> 68.12 | No clear effect |
| add TMI regularizer | TET VGGSNN · CIFAR-100, natural accuracy | accuracy (%): 72.66 -> 72.32 | Hurt |
| add output-entropy regularizer (H) instead of TMI | STBP VGGSNN · CIFAR-100, FGSM | accuracy (%): 13.51 -> 12.14 | Hurt |
| estimate MI with histogram (AIMIE) instead of KDE | STBP VGGSNN · CIFAR-100, FGSM | accuracy (%): 13.51 -> 12.54 | Hurt |
| use max-pooling over channels instead of averaging in TMI (ablation) | TMI regularizer · CIFAR-10 VGGSNN, FGSM | accuracy (%): 12.23 -> 25.80 | Improved |
| compute MI on the last spike layer rather than a shallow layer (ablation) | TMI regularizer · CIFAR-10 VGGSNN, FGSM | accuracy (%): 22.60 -> 25.80 | Improved |
| apply TMI temporal regularizer vs spatial MINE regularizer in ANN | VGGSNN vs ANN variant · CIFAR-100, FGSM | robustness improvement: 0.43 -> 1.69 | Improved |
| add TMI regularizer | AlexNet STBP · CIFAR-10, PGD TASA difference at last layer | TASA difference: 0.1661 -> 0.0802 | Improved |
| add TMI regularizer | SNN VGG11 T=4 · CIFAR-10, PGD | accuracy (%): 0.35 -> 26.87 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FEEL-SNN — FEEL-SNN: Robust Spiking Neural Networks with Frequency Encoding and Evolutionary Leak Factor (NeurIPS 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Robust Stable Spiking Neural Networks (ICML 2024 · similarity 0.90)
- Temporal Flexibility in Spiking Neural Networks: Towards Generalization Across Time Steps and Deployment Friendliness (ICLR 2025 · similarity 0.89)
- Enhancing the Robustness of Spiking Neural Networks with Stochastic Gating Mechanisms (AAAI 2024 · similarity 0.89)
- Enhancing Adversarial Robustness in SNNs with Sparse Gradients (ICML 2024 · similarity 0.89)
- FEEL-SNN: Robust Spiking Neural Networks with Frequency Encoding and Evolutionary Leak Factor (NeurIPS 2024 · similarity 0.89)
- Rethinking Spiking Neural Networks from an Ensemble Learning Perspective (ICLR 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.
