The Basics of GenAI: Theory and practice

This training is designed to provide you with a deeper insight into the fundamental workings of Large Language Models, which form the basis of many seminal Generative AI applications such as ChatGPT.

The training consists of a morning lecture and afternoon workshop:

10:00-12:00 ‘Understanding the basics of Generative AI’
This 2-hour lecture outlines the fundamental algorithmic structure of Large Language Models to clarify both the impressive capacities and remaining limitations of these models. It highlights why small changes in prompts can cause substantial differences in model behavior, and how LLMs can be used both with randomness or deterministically.

13:00-15:30 ‘Best practices for using Generative AI for academic research’
This workshop utilizes the insights of the morning lecture to teach best practices for using Generative AI to support academic research. Going beyond prompt engineering, it highlights effective workflows which consider both the basic limitations of LLMs as well as their versatile output formats.
Practical examples include data conversion, code generation, and various Natural Language Processing tasks such as summarization, literature review, and academic rewriting.

Goals

• Fundamental insight into the inner workings of Large Language Models
• Logical understanding of specific LLM phenomena such as ‘hallucinations’
• More concrete knowledge on how LLMs are trained and further refined
• Best practices for using Generative AI to support academic research
• Concrete examples of different parts of academic research workflows which involve using LLMs
• Hands-on experience with data conversion and code generation using LLMs
  

For this workshop you need to bring your laptop.

Dates and location : 

Teacher:

dr. Pieter Fivez (TEXTUA)

Language of the course:

English

Docop-points:

This course counts for 0,5 docop-points.

Registration:

 Via Sisa Self Service. Log in with your student account. Registration is possible as from 9 September 2025 and after you've been (re-)enrolled as a PhD student for 2025-2026.