More than 20 FAU students, mostly from the bachelor's program in Artificial Intelligence, joined the “Machine Learning and Data Analytics” Lab for the excursion. For about three hours, they gained an insight into the research that MPL scientists are conducting in the field of AI, and more precisely at the interface between physics, medicine, and AI.

Prof. Florian Marquardt welcomed the group and gave a general overview of the MPL, the Max-Planck-Zentrum für Physik und Medizin (MPZPM) and their research focuses. Under the title “AI for Physics and Physics for AI”, he explained how machine learning is used in physics to solve complex problems and how physical principles and methods can be used to increase the efficiency of machine learning. Prof. Marquardt discussed whether AI can comprehend the complexity of scientific thinking and what advantages this offers, for example for the construction of novel physical measuring instruments. Finally, he described how approaches from modern physics could advance artificial intelligence by replacing classical digital neural networks with hardware based on optical physics.

Prof. Marquardt then handed over to Timon Citak, a PhD student in Dr. Jona Kayser’s research group “Cellular Evolution”. In his talk “Machine Learning for Spatio-Temporal Evolution Prediction”, he showed how physics-based computer simulations using AI (including reinforcement learning) can be used to generate insights, which could be used to develop targeted cancer therapies. The goal of his research is to develop new evolution-based treatment strategies.

Last but not least, Dr. Carla Rodríguez from Dr. Mario Krenn’s research group “Artificial Scientist Lab” presented the XLuminA project. XLuminA is an AI framework which independently discovers new experimental designs in microscopy. Traditionally, the development of new microscopy techniques has relied on human intuition, which is challenging given the large number of possible configurations. XLuminA efficiently explores these possibilities. The framework has not only rediscovered fundamental microscopy techniques such as STED microscopy, but has also found new, powerful designs. XLuminA’s modular structure allows for flexible and versatile adaptations, thus promoting interdisciplinary collaboration.

The visitors were very interested and asked many questions after each of the three talks, which led to lively discussions between the students and the researchers. At the end of the visit, students from the International Max Planck Research School (IMPRS) held a small poster session for the FAU guests. In an informal atmosphere, they presented their work and provided information about the study programmes at the MPL.

Photos: © Elisabeth Offial