Fig. 1
From: Explaining and avoiding failure modes in goal-directed generation of small molecules
Experimental setup described by Renz and al. [12]. The initial dataset is split in two sets. The first split is used as a training set for the optimization model and the model-control model, and the second split for the data-control model. For a given molecule, the optimization (resp. model-control, data-control) score \(S_{opt}\) (resp. \(S_{mc}\), \(S_{dc}\)) is given by the optimization model’s (resp. model-control model’s, data-control model’s) predicted probability of being active The optimization score is used to guide goal-directed generation, and the evolution of control scores is also tracked during optimization. While the optimization score \(S_{opt}\) grows throughout training, the control scores \(S_{mc}\) and \(S_{dc}\) stagnates and reaches much lower values