Reproduce “Adversarial Attack and Defense for Non-Parametric Two-Sample Tests” 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

Xilie Xu, Jingfeng Zhang, Feng Liu, Masashi Sugiyama, Mohan Kankanhalli

The authors show that an adversary can make genuinely different sample sets appear alike to a two-sample test by perturbing one set to lower the test’s criterion. They train the deep-kernel test MMD-RoD to resist such changes. On CIFAR-10, a six-test ensemble’s power falls from 1.000±0.000 to 0.429±0.005; defense results vary by dataset and attack.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.