Catherine OlssonUnsolved research problems vs. real-world threat modelsAdversarial examples are worth studying, but most of the justifications for why exactly they’re worrisome strike me as overly literalMar 26, 2019Mar 26, 2019
Catherine OlssoninSouth Park CommonsHow to make your data and models interpretable by learning from cognitive science(This post accompanies a talk that Been Kim gave at South Park Commons on interpretable machine learning. If you want the technical details…Oct 12, 20171Oct 12, 20171