The research groups of Prof. I. Bogunovic and Prof. A. Lucchi at the Department of Mathematics and Computer Science at the University of Basel are inviting applications for a PhD position focused on the foundations and applications of reasoning with large language models (LLMs), including emerging diffusion-based LLM architectures. The successful candidate will investigate both theoretical aspects – such as understanding the mechanisms and limitations of reasoning in modern LLMs – and practical applications, including solving mathematical reasoning problems. The ideal candidate has a strong background in machine learning and an interest in bridging rigorous theoretical insights with challenging real-world tasks. They will also explore reinforcement learning strategies to optimize decision-making policies in complex environments, and develop fine-tuning protocols for large pre-trained models to adapt them to specialized reasoning domains as well as to address a generalization gap in LLM reasoning and their adaptation across reasoning domains. This work will blend cutting-edge experimentation – spanning RL, few-shot learning, meta-learning, etc. – with formal analysis to push the boundaries of what modern AI systems can reliably reason about.
Details and application links available here:
You are encouraged to send your application before April 20, 2026. However, there is no deadline for applying, and the position will be open until a suitable candidate is found.