SmallData Seminar: “Bring Your Cool Tool” Highlights Cutting-Edge Methodologies

The first SmallData Seminar titled “Bring Your Cool Tool” successfully showcased innovative methodological tools being utilized in current research projects.

Our Doctoral Researcher, Fabian Kabus (Fusion Hub) kicked off the seminar with an insightful introduction to PyTorch, covering essential topics such as tensor operations, dynamic computation graphs, Autograd, evaluation mode, and gradients. His session provided a solid foundation for understanding this powerful deep-learning framework.

Fabian also teamed up with our Doctoral Researcher Simon Ging (B05) to provide a presentation on “AI-Based Coding with GitHub Copilot.” They demonstrated the installation process and showcased how Copilot can provide real-time feedback and suggestions as you code, making it an invaluable tool for developers.

Our Clinician Scientist, Philipp Arnold (B05), concluded the seminar with a presentation on large language models (LLMs). He explained the fundamentals of LLMs, their operational mechanisms, and provided a live demonstration of various applications. Philipp also discussed the advantages of hosted versus local LLMs, emphasizing the use of local models for handling sensitive data, and demonstrated use cases such as extracting structured text from radiology reports.

We give a big thanks to our SmallData presenters, and of course, to all the participants for attending and putting forward insightful questions.

If you would like to hear about our other events, subscribe to our mailing list!

Administrative Manager

Marc Schumacher

Institute of Medical Biometry and Statistics,
Faculty of Medicine and Medical Center –
University of Freiburg