Retrieval Augmentation Generation (RAG) has emerged as one strategy to get more accurate and relevant information on a specific topic while maintaining the versatility of generative AI. In this seminar, three researchers - Carsten Bergenholtz, Oana Vuculescu and Niels Lauritzen - will take the participants through setting up a RAG system for educational purposes.

Program:

·  Introduction to RAG: Overview of Retrieval Augmented Generation, highlighting how it complements and differs from standalone Large Language Models.

·  System Showcases: Examples of RAG in action in a Math, Philosophy of Science and Organizational behavior class – for problem solving, explaining concepts and providing feedback.

·  Live RAG Setup: Demonstration of setting up a RAG system using audience-provided documents, with comparisons to commercially available tools like Open AI GPTs or Gemini Gems.

·  Q&A and Getting Started: An interactive session where participants can ask questions and get guidance on what it takes to set up a RAG system or how to configure GPTs or Gem-bots.

Location: 1481-237.

Date and time: January 29, 10.00-15.00

The seminar is free and open to all Aarhus University employees.