Evaluating Explainable AI: Which Algorithmic Explanations Help Users Predict Model Behavior?

ACL 2020