PhD Positions in Reinforcement Learning for LLMs and Agentic AI (Prof. I. Bogunovic) University of Basel in Switzerland

The Rhine AI (Reasoning, Human-aligned Intelligence and Novel Exploration) group, led by Prof. Ilija Bogunovic at the Department of Mathematics and Computer Science, University of Basel, is seeking multiple PhD students to join their research on reinforcement learning for large language models (LLMs) and agentic AI. The group focuses on developing next-generation post-training algorithms, exploring diffusion-based reasoning with language models, aligning AI systems with complex human values, and building self-improving agents capable of autonomous learning. Research in the group combines advanced experimentation—including reinforcement learning, meta-learning, and robust optimization—with rigorous theoretical analysis. The group publishes in top venues such as ICML, NeurIPS, ICLR, JMLR, and AISTATS, and collaborates with national and international AI and ML researchers. The ideal candidate will have a strong background in machine learning and artificial intelligence, and be comfortable with or eager to learn large-scale multi-GPU experimentation on challenging LLM tasks. Applicants must hold a Master's degree in computer science, mathematics, or electrical engineering, with a strong academic record and mathematical background. Experience with large-scale computational experiments is advantageous, but not required. Excellent communication skills and fluency in English are essential. The group is committed to increasing diversity and encourages applications from women and other underrepresented groups. The position is based in Basel, Switzerland, and offers close supervision, access to international collaborators, a state-of-the-art workstation with NVIDIA H200 GPUs, a dynamic research environment, opportunities for international conference travel, and excellent salary and benefits. The PhD duration is four years, with a flexible start date around January 2026. Applications should be submitted as a single PDF including a motivation statement, CV, transcripts, contact information for recommenders, and a thesis or publication. Early applications are encouraged, but the position will remain open until filled. Short-listed candidates will be contacted within 2-3 weeks of application. Due to high application volume, only those selected for interview will be notified.