Interpret is an interpretable machine learning library and glassbox model framework. It provides toolkits for training inherently transparent models and applying post-hoc explanation techniques to make machine learning predictions human-understandable.
The framework distinguishes itself by integrating differential privacy into the training of interpretable models to prevent sensitive data from leaking through explanations. It also features a visualization tool for rendering interactive decision paths and model behavior.
The library covers model explainability through feature importance calculation, interaction detection, and the generation of local and global explanations. It includes capabilities for auditing models via JSON serialization, enforcing monotonicity constraints, and approximating black-box systems.
The toolkit supports model management utilities such as model aggregation, merging, and the editing of trained model components.