The successful candidate will develop generative machine-learning methods for amorphous molecular thin films - the supramolecular structures that govern the performance of organic-electronic materials. Equilibrating such films by brute-force molecular dynamics is prohibitively slow. Instead, we build them using diffusion models at coarse-grained resolution, formulated so that the model yields not only realistic structures but also free energies. The work sits at the interface of statistical mechanics, molecular simulation, and deep generative modeling, and is embedded in SIMPLAIX, a research initiative on multiscale simulation and machine learning funded by the Klaus Tschira Foundation.

Profile:

• Solid grounding in statistical mechanics
• Experience with machine learning and/or molecular simulation; strong Python and PyTorch skills
• Genuine interest in method development, and comfort with mathematical formalism
• A degree in physics is required for admission to the Heidelberg Graduate School for Physics (HGSFP). This is a formal admission requirement.

Conditions:

The position is funded for three years at 75% of salary group E13 TV-L. Please note that a 75% appointment is the standard arrangement for doctoral researchers in German physics departments: this is a full doctoral position, and the remaining time is your own thesis work. It is not a part-time job requiring additional income.

The working language of the group is English; no German is required.

Applications are reviewed on a rolling basis until the position is filled. If the deadline shown above has passed, please check the group website or write directly - the position may still be open.

Full details: https://tristanbereau.com/positions/gen_ml_morphologies.html