Research map / Adversarial robustness and security / Certified and empirical robustness
Toward Robust Spiking Neural Network Against Adversarial Perturbation
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 |
|---|---|---|---|
| apply S-IBP/S-CROWN certified robust training to the SNN | CONV SNN · MNIST untargeted white-box attack | attack error: 50.7% -> 13.0% (-37.7%) | Helped |
| apply S-IBP/S-CROWN certified robust training to the SNN | CONV SNN · NMNIST untargeted white-box attack | attack error: 49.0% -> 12.0% (-37.0%) | Helped |
| apply S-IBP/S-CROWN certified robust training to the SNN | FC SNN · FMNIST untargeted white-box attack | attack error: 49.7% -> 38.7% (-11.0%) | Helped |
| apply robust training at small attack epsilon | CONV SNN · FMNIST attack eps 0.040 | attack error: 19.7% -> 22.0% (+0.3%) | Hurt |
| apply robust training | SNN · MNIST/FMNIST/NMNIST clean error | original error: 0.3% -> 4.0% (MNIST CONV) | Hurt |
| use robust training time steps T'=3 instead of other values (1 or 5) (ablation) | S-CROWN robust training · MNIST CONV eps 0.12 | original accuracy and S-CROWN loss | Helped |
| increase robust training time steps T' from 1 to 3 to 5 (ablation) | S-CROWN robust training · training time per epoch | relative training time: 2.6x, 7.7x, 12.9x | Hurt |
| use CONV instead of FC network for robust training | S-CROWN robust training · MNIST attack | attack error reduction: -37.7% (CONV) vs -25.7% (FC) | Helped |
Closest papers by meaning
- Enhancing Adversarial Robustness in SNNs with Sparse Gradients (ICML 2024 · similarity 0.93)
- Robust Stable Spiking Neural Networks (ICML 2024 · similarity 0.92)
- FEEL-SNN: Robust Spiking Neural Networks with Frequency Encoding and Evolutionary Leak Factor (NeurIPS 2024 · similarity 0.90)
- Synergy Between the Strong and the Weak: Spiking Neural Networks are Inherently Self-Distillers (NeurIPS 2025 · similarity 0.89)
- Enhancing the Robustness of Spiking Neural Networks with Stochastic Gating Mechanisms (AAAI 2024 · similarity 0.88)
- EnOF-SNN: Training Accurate Spiking Neural Networks via Enhancing the Output Feature (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.