Unpacking Black-Box Models: ExSum Mathematical Framework To Evaluate Explanations of Machine-Learning Models

MIT researchers create a mathematical framework to evaluate explanations of machine-learning models and quantify how wel