The interpretability of a machine learning model’s inner workings