FastSVERL provides a practical and scalable approach for principled, rigourous interpretability in reinforcement learning.

We combine Monte Carlo and regression-based methods to get a flexible estimator which achieves state-of-the-art performance.

A complexity-theoretic analysis of exact, tractable SHAP explanations for Tensor Networks and their implications for other ML models

SHAP zero estimates Shapley values and interactions with a near-zero marginal cost for future queried sequences through the Fourier transform.

The introduced TreeHFD algorithm estimates the Hoeffding or ANOVA functional decomposition of tree ensembles from a data sample, and therefore provides an efficient explainability method through a set of low dimensional functions.