Welcome to Agentic Large Language Models at LIACS, Leiden University!

About This Course

This course explores agentic LLMs — LLMs that reason, act, and interact — following our course textbook Agentic Large Language Models, written by the teaching team: Max van Duijn, Michiel van der Meer, Aske Plaat, and Niki van Stein. Lectures start Tuesday, September 1, 2026.

What to Expect

Throughout the course, you will:

  • Learn the fundamentals of Large Language Models: tokenization, transformers, training pipelines, and scaling laws
  • Study advanced reasoning techniques: chain-of-thought prompting, self-reflection, and world models
  • Build your own agentic LLMs across four hands-on case studies, from training a model from scratch to running a multi-agent social simulation
  • Master tool use, retrieval augmentation (RAG), and the Model Context Protocol (MCP)
  • Explore multi-agent systems and the behavioral and cognitive perspectives on agentic LLMs

Getting Started

  1. Review the Syllabus — understand the course structure, topics, and assessment
  2. Check the Schedule — see what we’ll cover each week, and when each of the four assignments is due
  3. Explore Resources — find the papers, frameworks, and benchmarks used in the book
  4. Set up your environment — make sure you have Python and PyTorch installed and can access LLM APIs

Stay Connected

Make sure to:

  • Attend all lectures
  • Start early on the case-study assignments — each spans about three weeks
  • Reach out to the teaching team if you have questions

We’re excited to have you in the course. Let’s build some amazing agentic LLMs together!

Posted by the course team