Feature Learning for Interpretable, Performant Decision Trees Supplementary Material 1 Experiment Specification