This is a subset of the dataset used to train general-purpose CH GAP interatomic potential [1]. This subset contains the interacting molecules generated stochastically in a following manner. The subset is generated using active learning and uncertainty-based configuration selection. We started with randomly chosen pairs of CH-containing molecules from the QM9 database up to 7 carbon atoms. The probability of selecting the molecules is set based on the energy and size of the molecule. The probability is lower as the energy above the convex hull is higher. A bigger size of the molecule also lowers the probability to favor the inclusion of small structures. Then, we estimate the uncertainties for the new structures based on how far away from the existing interacting molecules in the training set they are (in configuration space), and identify those with the largest expected errors. This way, we generated about 3k structures.

2024