Bridging CGR Representations and Language Models for Reaction Property Prediction
This research presents a new generation of graph-based models that provide real-time prediction of reaction properties, such as activation energies, enthalpies, rate constants and yields, in relation to discrete-time reactions.