AI4DH Lecture: Understanding Large Language Models for Digital Humanities

On Monday, 2 February 2026, the second lecture for digital humanities and social science scholars was held, titled “Understanding Large Language Models for Digital Humanities”.

Prof. Dr Marko Robnik-Šikonja presented the workings of the transformer architecture in neural networks and decoder models, which are used in generative models such as ChatGPT. The lecture explained the construction of decoders, pretraining, instruction following, preference alignment, and fine-tuning. The final part focused on prompting strategies, including in-context learning and chain-of-thought reasoning.

This seminar provides the foundational knowledge for the next seminar, “Adapting and Fine-tuning LLMs – a Hands-on Approach for DH”.