Explainable Artificial Intelligence via Representer Points & Infidelities