Reproduce “Practical Auditing of Differentially Private Graph Release Mechanisms: Methodology, Toolkit, and Case Study” 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
Nicholas D'Silva, Surya Nepal, Salil S. Kanhere
The authors present an empirical auditing toolkit for testing differential privacy claims in graph-release mechanisms. It pairs meaningful neighboring graphs, called canaries, with scalar witness statistics and statistical certification. Auditing PrivDPR, edge count yields a 99%-confidence lower bound of ε = 2.053 against claimed ε = 0.1 at δ = 10⁻⁵ for one connected pair, not a universal privacy estimate.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.