Would you like to work in an innovative and international setting? Then MCL is the perfect work place for you!
Your Profile
You have—or will shortly receive—a doctoral degree in any of the computational sciences (e.g., computational materials science/physics/ chemistry, applied mathematics). The project requires knowledge in
- linear algebra/machine-Learningmethods,
- programming (e.g., C/C++, Python),
- crystallography/atomic-scale mechanics,
- atomistic simulation,
and you are enthusiastic about mastering them to develop new AI methods to accelerate the design of innovative materials.
Tasks & Responsibilities
Machine-learning interatomic potentials (MLIPs) have been a breakthrough for providing the quantitative accuracy of quantum mechanics to atomistic simulations. However, they suffer from a very high complexity (e.g., millions of parameters) when applied to systems with a large chemical space, also involving other quantummechanical phenomena like magnetism.
In this project, you will explore tensor networks, a way of compressing high-dimensional parameter tensors with several lower-order tensors, as a representation of machine-learned atomic descriptors. The challenge of applying tensor networks to MLIPs are symmetries (e.g., rotational invariance of atomic energies), which need to be preserved by the tensor network. Your main tasks will revolve around advancing compression algorithms to such symmetrypreserving tensor networks.
Our Offer
- flexible working hours
- further training/education
- home office
- employee events
- cooperation with Firmenradl
A permanent employment contract starting from January 2026 and a gross salary per year of € 64.618,40 (40h/week). The above project on which you are being hired is planned for a duration of 2.5 years.
Please send us your application and a detailed resume. We are looking forward to it! We would like to especially encourage women to apply.