Reliable Adversarial Robustness Evaluation for Spiking Neural Networks
Research Paper#Spiking Neural Networks, Adversarial Robustness, Machine Learning🔬 Research|Analyzed: Jan 3, 2026 16:26•
Published: Dec 27, 2025 08:43
•1 min read
•ArXivAnalysis
This paper addresses the challenge of evaluating the adversarial robustness of Spiking Neural Networks (SNNs). The discontinuous nature of SNNs makes gradient-based adversarial attacks unreliable. The authors propose a new framework with an Adaptive Sharpness Surrogate Gradient (ASSG) and a Stable Adaptive Projected Gradient Descent (SA-PGD) attack to improve the accuracy and stability of adversarial robustness evaluation. The findings suggest that current SNN robustness is overestimated, highlighting the need for better training methods.
Key Takeaways
- •Proposes a more reliable framework for evaluating SNN adversarial robustness.
- •Introduces Adaptive Sharpness Surrogate Gradient (ASSG) to improve gradient accuracy.
- •Designs Stable Adaptive Projected Gradient Descent (SA-PGD) for faster and more stable convergence.
- •Demonstrates that current SNN robustness is overestimated.
- •Highlights the need for more dependable adversarial training methods.
Reference / Citation
View Original"The experimental results further reveal that the robustness of current SNNs has been significantly overestimated and highlighting the need for more dependable adversarial training methods."