Scrim in ML, known as Soft Classifiers with a Rules Interpretable Margin, is a technique used in machine learning that aims to improve model interpretability by incorporating