Our paper “Foundations of Symbolic Languages for Model Interpretability” got accepted as spotlight for NeurIPS 2021. Check it out here.