PhD Position in Post-Training of Diffusion Language Models (Prof. I. Bogunovic)

Universität Basel; Basel; On-site; CHF 47,000 - 57,000; Full time

Benefits offered by this job

Close supervision

Conference travel

Dynamic environment

Salary & benefits

Job summary

Universität Basel invites applications for a PhD position funded by the Swiss National Science Foundation (SNSF) project. The role focuses on post-training methods for diffusion-based language models, with emphasis on reasoning and evaluation benchmarks.

The successful candidate will develop reinforcement learning approaches, explore reasoning-aware generation, and contribute to benchmarking across domains.

Qualifications

• Master's degree in a related field with strong academic record.
• Solid mathematical foundation and ability to run large-scale experiments.
• Experience with LLMs, reinforcement learning, or generative modeling is a plus.
• Research experience is advantageous but not mandatory.

Responsibilities

• Conduct research in machine learning and artificial intelligence.
• Write research papers and present results at leading international conferences.
• Interact with national and international teams of AI and machine learning experts.

Skills

Machine learning background

Coding skills

LLMs / RL / Generative modeling

Education

Master's degree in computer science, mathematics, physics

Job description

The research group Rhine AI of Prof. I. Bogunovic at the Department of Mathematics and Computer Science at the University of Basel is inviting applications for a PhD position funded by the Swiss National Science Foundation (SNSF) project "Foundations of Post-Training for Reasoning in Diffusion Large Language Models." The position focuses on post-training large diffusion-based language models (dLLMs), an emerging alternative to standard autoregressive LLMs, to improve their reasoning capabilities. The successful candidate will develop reinforcement learning methods for post-training large dLLMs, explore reasoning-aware generation strategies such as adaptive unmasking and credit assignment, and help build benchmarks for evaluating reasoning across domains, including mathematical problem solving. The ideal candidate has a strong background in machine learning and is excited about large-scale experimentation with modern AI systems, supported by principled analysis where it helps. This work will push the boundaries of what modern AI systems can reliably reason about. The position is available starting January 2027 or later. The duration of a PhD in Switzerland is 4 years.

Your position

• Conduct research in machine learning and artificial intelligence
• Write research papers and articles, and present results at leading international conferences. We publish in venues such as ICML, NeurIPS, and ICLR
• Interact with national and international teams of AI and machine learning experts

Your profile

Applicants must possess a Master's degree in computer science, mathematics, physics, or a related field. Candidates should have an exceptional academic record and a solid mathematical foundation. They are also expected to have strong coding and implementation skills, with the ability to run large-scale experiments. Experience with LLMs, reinforcement learning, or generative modeling is a plus. Research experience is advantageous but not mandatory.

Excellent communication skills and fluency in English (spoken and written) are required.

We also seek to increase the number of women and other groups in areas where they are underrepresented and therefore explicitly encourage such groups to apply.

We offer you

• Close supervision and mentoring, working with and access to national (EPFL and ETH among others) and international collaborators
• A diverse range of tasks in a dynamic environment
• The possibility of traveling to international conferences to meet distinguished scientists
• Excellent salary, benefits, and working conditions