We hire students from relevant programs at UdS (e.g., LST, CS) as student assistants. Student assistants in our lab are closely involved in research and often even lead their own projects and have published strong papers (e.g., at ACL 2024, NeurIPS 2024, COLM 2025, NeurIPS 2025). Please note that we expect applicants to have excellent programming skills, and strong grades in a relevant technical field (e.g., grades close to 1,0 in the German academic system; or strong GPAs from other countries).
We offer fully funded PhD positions. As is typical in Europe, PhDs are expected to take 3-4 years. Our PhD students are affiliated with the Saarbrücken Graduate School of Computer Science.
Our lab is generously funded by a prestigious Emmy Noether grant from the German Research Foundation. We have plenty of compute resources by academic standards.
Applicants who will have an MSc by the start date of the PhD: Please directly apply to our lab by email. Admitted students will directly join our lab. Start dates are flexible and can be negotiated. In your email to Michael (mhahn@lst.uni-saarland.de), please include:
For the current round of recruitment, applications received by 20 December 2025 will receive full consideration. For your reference, you can find the legally binding call for the position here.
Aplicants without an MSc: Applicants who only have an undergraduate degree should instead apply to the Saarbrücken Graduate School of Computer Science, in which case the PhD includes an initial preparatory phase. If you’re interested in working with me, please send me an email to confirm in advance.
The application process is the same as for PhD students.
We take interns (undergraduate or graduate) around the year. If you’re an undergraduate from outside Germany, we may be able to provide housing but generally can’t pay a stipend. If you’re a graduate student, we might also be able to pay you.
As a member of our group, you’d have substantial freedom in choosing your research topic. We are mostly interested in understanding the abilities, limitations, and inner workings of language models. You can take a look at our recent papers in NeurIPS, ICLR, ICML, and COLM to get a feel for the kind of work you might be doing with us.
If you’re interested in joining, please send your application to Michael (mhahn@lst.uni-saarland.de).