BO-ICL

BO-ICL is a tool that utilizes Bayesian optimization and in-context learning to predict molecular properties and generate new molecular data. It allows users to perform regression with uncertainties and supports inverse design for proposing new molecular structures based on desired properties.