Michael Petrowski

Hi, I’m Michael. I am currently pursuing an MSc in Artificial Intelligence and Data Science at Heinrich Heine University Düsseldorf. My passion for research was ignited during my Bachelor’s thesis, leading me to my current role as a Student Researcher at the Dialog Systems and Machine Learning Lab under Prof. Milica Gašić.

My thesis investigated whether equipping artificial agents with introspection could enhance adaptive learning. I developed a ‘self-aware’ reinforcement learning framework where agents infer internal ‘pain’ states to dynamically guide exploration. This approach demonstrated that self-awareness significantly improves performance while replicating complex, human-like behavioral dynamics. The resulting paper, Exploration Through Introspection, has been accepted at the AAAI-26 Workshop on Theory of Mind. Currently, I am expanding this research by incorporating this introspective reward framework into Large Language Models to optimize complex task-oriented dialogue systems.