CRC 1114 aims at methodological developments for the modelling and computational simulation of complex processes involving many (more than two) interacting scales, driven by real-life applications from the bio-, geo-, and material sciences. In its individual projects, mathematicians cooperate with colleagues from the natural sciences to advance both their modelling capabilities and their insight into concrete multiscale phenomena.

Job description:
This PostDoc position is part of the CRC1114 project A04 (Efficient calculation of slow and stationary scales in molecular dynamics) and is situated within the Al4Science group (Prof. Noé) at Freie Universität Berlin. The group develops new machine learning techniques for the molecular sciences. Aim of the project is the development of new theory and machine learning methods for the coarse-graining of quantum or classical many-body systems. These methods will be of importance for the development of drugs, materials as well as fundamental research in chemistry and biology.

Requirements:
• BSc or MSc in Physics, Chemistry, Computer Science or related areas.
• Completed PhD, ideally with focus on method development in quantum Chemistry or statistical Mechanics.

Desirable:
• Experience in the development of machine learning systems and neural networks, especially Graph Neural Networks, Transformer, invariance and equivariance, statistical learning theory.
• Experience in the modeling and simulation of molecular systems.
• Extensive programming experience in Python, and with PyTorch or JAX.
• Scientific publications.
• Fluent in written and spoken English.

For further information, please contact Prof. Dr. Frank Noé (frank.noe@fu-berlin.de).