The Cybersecurity and AI Group at Freie Universität Berlin, Department of Mathematics and Computer Science, is directed by Prof. Dr.-Ing. habil. Gerhard Wunder and supported by Bundesdruckerei GmbH. Main working areas are AI, Deep Learning and Information Theory for Security, Privacy and Identity. You can find further details at: www.mi.fu-berlin.de/en/inf/groups/ag-comm/index.html.

Job description:
The C-AI group seeks an outstanding PhD candidate in the field of trustworthy Large Language Models (LLMs) with a strong focus on Explainability, Causality and Watermarking of LLMs. The position is part of the new BMBF initiative “Chances and Risks of Generative AI” initiated by the Bundesministerium für Bildung und Forschung led by Fraunhofer AISEC and carried out together with CISPA Helmholtz Center for Information Security and TU Berlin (until end 2026). The PhD candidates contributes to the theoretical foundations, design, and implementation of Trustworthy LLMs thereby closely collaborating with the partners in the consortium. Candidates are expected to co-author articles published in high-quality academic venues such as IEEE/ACM conferences and journals. Within the framework of the externally funded research project, the opportunity for writing a doctoral-thesis is granted.

Requirements:
Applicants must possess a completed scientific degree (master or equivalent) in computer science, mathematics, electrical engineering or similar.

Desirable:
Profound theoretical knowledge in AI, Deep Learning and good coding skills in either Python, TensorFlow (or PyTorch) as well as project management tools (Git etc.). Experience in the implementation and training of LLMs as well as Prompt Engineering and RAG is a plus. Ability and willingness to work and cooperate with the members of group and the project consortium will be valued.