Reproduce “Learning to Localize Leakage of Cryptographic Sensitive Variables” 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
Jimmy Gammell, Anand Raghunathan, Abolfazl Hashemi, Kaushik Roy
ALL ranks timepoints in cryptographic power or electromagnetic traces by usefulness for inferring a secret. It trains a predictor alongside a budget-limited mask that hides measurements, then uses learned occlusion probabilities as a leakage ranking. In a constructed pair-only test, ALL succeeds at pair counts up to 64 times higher than prior deep-learning approaches; this does not establish hardware performance.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.