About Me
I am a Ph.D. candidate in the research group of Prof. Dr. J. Marius Zöllner. In my capacity as a Teaching Assistant for Machine Learning 1 (Fundamentals) and Machine Learning 2 (Advanced Methods), I support students in mastering both the theoretical foundations and the implementation of state-of-the-art ML techniques.
Research Interests
The central challenge in building generalist agents, systems that can perform a wide range of tasks without being individually trained for each one, is data. Current approaches depend on large quantities of task-specific, action-labeled demonstrations, a pipeline too expensive and narrow to scale to general competence. My research asks whether this bottleneck can be bypassed by learning from the web. I develop methods that extract implicit behavioral knowledge from unannotated video corpora, building the foundation for generalist policies that acquire broadly transferable skills directly from observing how humans and other embodied systems interact with the world. The web offers an extraordinary window into human behavior — cooking tutorials, assembly guides, surgical procedures, sports footage — billions of hours of people manipulating objects, navigating environments, and solving physical tasks. If we can learn from this data, the bottleneck dissolves and we get closer to generalist agents.
Collaboration & Supervision
I am actively seeking motivated students for Hiwi positions or Thesis projects (M.Sc.). If you are interested in these research areas, please reach out via email to discuss potential collaborations.
#MachineLearning #AutonomousAgents #FoundationModels #ImitationLearning #LargeScale #Robotics
Open Positions
- Student Research Assistant Position in Generally Capable Autonomous Agents from Web-Scale Video Data
Aktive Projekte
 |
SofDCar |
Forschungsgebiete
- Maschinelles Lernen, Deep Learning, Künstliche Intelligenz