Dear all,
We are hiring a fully funded PhD student at Linköping University (LiU), Sweden, on generative modeling and synthetic data for data-efficient machine learning, with fairness and (differential) privacy guarantees.
The project combines generative deep learning (diffusion models, representation-conditioned generation) with coreset theory to construct small synthetic data summaries that let standard learning algorithms produce trustworthy models, moving the approximation from the model to the data. The work spans low-dimensional classical models to high-dimensional deep generative models.
The position is part of the Wallenberg AI, Autonomous Systems and Software Program (WASP,
https://wasp-sweden.org), including a structured graduate school, an industrial partner network, and a potential semester-long international research visit.
Why this position is attractive for international applicants:
- We regularly publish at top machine learning venues.
- Fully employed with monthly salary (not a stipend), pension, healthcare, and paid parental leave; four years, extendable to five with teaching duties.
- No language barrier: English is the working language at the university, and Swedes generally speak English very well.
- State-of-the-art computing: access to Berzelius (Sweden's dedicated AI supercomputer, hosted at LiU), and LiU is hosting the upcoming EuroHPC Arrhenius and the MIMER AI Factory.
- Sweden offers a high quality of life, flat academic hierarchies, five weeks of vacation, and easy access to nature.
Supervisors: Gabriel Eilertsen (Division of Media and Information Technology, Campus Norrköping) and Sebastian Mair (Division of Statistics and Machine Learning, Campus Valla in Linköping). Joint affiliation with both divisions, with a desk at both campuses.
Deadline: August 21, 2026
Full ad and application:
https://liu.se/en/work-at-liu/vacancies/29453Questions:
sebasti...@liu.sePlease help spread the word to interested candidates.
Best,
Sebastian