3 papers across 2 sessions
This study proposes a Bayesian Flow Network integrated with gradient-based guidance to generate molecules satisfying diverse properties.
A novel method for fast 3D molecule generation