Research map / Adversarial robustness and security / Certified and empirical robustness
SNN-RAT: Robustness-enhanced Spiking Neural Network through Regularized Adversarial Training
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 |
|---|---|---|---|
| add spectral/orthogonal weight regularization (REG) (ablation) | SNN VGG-11 vanilla · CIFAR-10 FGSM(BPTT) | accuracy (%): 12.78 -> 26.60 | Helped |
| add mixed adversarial training (MIX) of FGSM/RFGSM with BPTT/BPTR (ablation) | SNN VGG-11 vanilla · CIFAR-10 FGSM(BPTT) | accuracy (%): 12.78 -> 37.49 | Helped |
| add REG on top of MIX (ablation) | SNN-RAT VGG-11 · CIFAR-10 PGD(BPTT) | accuracy (%): 16.63 -> 21.16 | Helped |
| add REG on top of MIX (ablation) | SNN-RAT VGG-11 · CIFAR-10 clean | accuracy (%): 92.04 -> 90.74 | Hurt |
| full RAT vs vanilla training | SNN VGG-11 · CIFAR-10 clean | accuracy (%): 93.06 -> 90.74 | Hurt |
| full RAT vs vanilla training under PGD(BPTT) white-box | SNN VGG-11 · CIFAR-10 | accuracy (%): 00.04 -> 21.16 | Helped |
| full RAT vs vanilla under PGD(BPTT) black-box | SNN VGG-11 · CIFAR-10 | accuracy (%): 04.61 -> 77.43 | Helped |
| use conversion-based approximation (CBA) instead of BPTT/BPTR for attack | gradient attacks on SNN VGG-11 · CIFAR-10 PGD | accuracy under attack (%): 37.30 vs 00.04 | Hurt |
| RAT vs prior robust SNN training (Kundu et al.) under FGSM | SNN VGG-11 · CIFAR-100 | accuracy (%): 22.00 -> 25.86 | Helped |
| orthogonal projection rate beta variation (ablation) | SNN-RAT · CIFAR-10 | clean accuracy | No clear effect |
Closest papers by meaning
- Enhancing Adversarial Robustness in SNNs with Sparse Gradients (ICML 2024 · similarity 0.95)
- Robust Stable Spiking Neural Networks (ICML 2024 · similarity 0.91)
- Towards Effective and Sparse Adversarial Attack on Spiking Neural Networks via Breaking Invisible Surrogate Gradients (CVPR 2025 · similarity 0.89)
- FEEL-SNN: Robust Spiking Neural Networks with Frequency Encoding and Evolutionary Leak Factor (NeurIPS 2024 · similarity 0.88)
- Enhancing the Robustness of Spiking Neural Networks with Stochastic Gating Mechanisms (AAAI 2024 · similarity 0.88)
- Synergy Between the Strong and the Weak: Spiking Neural Networks are Inherently Self-Distillers (NeurIPS 2025 · 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.