Machine-learning models can make mistakes and be difficult to use, so scientists have developed explanation methods to help users understand when and how they should trust a model's predictions. These ...