Supervising Team: Jakub Šimko (KInIT), Dominik Macko (KInIT)

Advances in large language models have enabled the generation of texts that are, at first sight, almost indistinguishable from human writing. Our work focuses on authorship attribution in a setting where human-written texts coexist with texts generated by different large language models. The objective is to design a solution capable of reliably distinguishing the origin of texts, with a strong focus on the transparency and explainability of the results.