Research
Samir Darouich develops generative models for computational chemistry, particularly molecular structures and chemical transition states. His work investigates how pretraining can improve predictions for unfamiliar chemical environments and how geometric symmetries can be incorporated into efficient generation methods. Recent work includes generating molecular conformations and transition states in a single step.
Selected publications
- SymDrift: One-Shot Generative Modeling under Symmetries 2026
- Robust generative transition-state models for unseen chemistry 2026
Information reviewed 26 September 2026 · University profile