<ul><li>This research explores a model for explainable AI that provides proof as explanations for reliable predictions.</li><li>The model defines an explanation as a subset of the training data that can serve as a proof of a prediction's correctness.</li><li>The research presents the concept of the robust hollow star number to determine the worst-case size of the smallest certificate achievable.</li><li>The study also analyzes worst-case distributional bounds and distribution-dependent bounds for certificate size.</li></ul>

Proofs as Explanations: Short Certificates for Reliable Predictions

Discover more