Prime Implicant Explanations for Reaction Feasibility Prediction
This work addresses interpretability for chemists and researchers in automated synthesis planning, but it is incremental as it focuses on small-scale tasks and builds on existing explanation methods.
The authors tackled the lack of transparency in machine learning models for predicting chemical reaction feasibility by introducing a novel formulation of prime implicant explanations tailored to this domain, with preliminary experiments showing that these explanations conservatively capture ground truth by consistently including essential molecular attributes.
Machine learning models that predict the feasibility of chemical reactions have become central to automated synthesis planning. Despite their predictive success, these models often lack transparency and interpretability. We introduce a novel formulation of prime implicant explanations--also known as minimally sufficient reasons--tailored to this domain, and propose an algorithm for computing such explanations in small-scale reaction prediction tasks. Preliminary experiments demonstrate that our notion of prime implicant explanations conservatively captures the ground truth explanations. That is, such explanations often contain redundant bonds and atoms but consistently capture the molecular attributes that are essential for predicting reaction feasibility.