Effective Prompting

In this workshop, you will learn how to write effective prompts for generative AI models like ChatGPT, Claude or Gemini.

We start by investigating how these AI models actually work: we explore how they were trained, what’s going on behind the scenes when we interact with them, and why they respond the way they do. We look at what tasks Large Language Models (LLMs) currently excel at, as well as areas where they still struggle.

Next, the hands-on part of the workshop gives an overview of key prompting strategies for the optimal use of LLMs. You will not only be introduced to basic techniques such as few-shot prompting, but also to more advanced methods such as chain-of-thought and chain-of-verification prompting, always based on the latest scientific insights. Moreover, you will work with GenAI tools yourself, to experience how you can teach their models new tasks, improve their reasoning abilities and reduce their tendency to hallucinate.

Finally, we wrap up with some important challenges that come with using generative AI, including ecological considerations, copyright and the risk of plagiarism.