Reproduce “Characterizing Privacy Leakage in Spiking Neural Networks: A Systematic Analysis of Temporal Dynamics, Stochasticity, Neuronal Dynamics and Learning Mechanisms” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ayana Moshruba, Maryam Parsa
The authors test how temporal activity in spiking neural networks affects membership leakage, where an attacker infers whether an example was used for training. They compare input encodings, neuron dynamics, and learning regimes using one black-box shadow-model attack. Poisson-Burst lowers Fashion-MNIST AUC from 0.534 to 0.502 versus Rate, but findings are limited to tested settings and attack.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.