team

Dr. Florian Mai leads the mAI alignment lab, focusing on AI alignment and safety research. His work spans value alignment, reasoning models, and the development of safe large scale AI systems.

Projects:

• Scalable Oversight by Learning to Decompose Tasks
• Emergent Misalignment
• Backdoor Detection
• Value Alignment
• Learning to Plan from Unlabeled Data

David is a PhD student at the University of Bonn under the supervision of Dr. Florian Mai and Prof. Dr. Lucie Flek since August 2024. His current research interests include emergent misalignment, reinforcement learning, and the training of large language models.

Projects:

• Emergent Misalignment

Lukasz is a research assistant at the University of Bonn. He received his PhD in theoretical physics from Adam Mickiewicz University in Poznań before transitioning to AI safety.

Projects:

• Scalable Oversight by Learning to Decompose Tasks

Robin is an undergraduate computer science student at Saarland University and a student assistant at the University of Bonn.

Projects:

• Backdoor Detection

Mohammad is a master’s student at the University of Bonn.

Projects:

• Emergent Misalignment

Sayantak was an MSc. student at the University of Bonn, working on learning to plan for large language models from unlabeled data.

Projects:

• Scalable Oversight by Learning to Decompose Tasks
• Learning to Plan from Unlabeled Data