The IberAuTexTification competition focuses on authorship identification in six languages of the Iberian Peninsula (Spanish, Catalan, Basque, Galician, Portuguese, and English) and spans seven domains (Chat, How-to, News, Literary, Reviews, Tweets, and Wikipedia). It includes two subtasks: MGT Detection, a binary classification task to determine whether a text was written by a human or a large language model (LLM), and MGT Attribution, a multiclass classification task in which participants must identify which of six LLMs generated a machine-generated text.
Forum
Year
2024
Link to publication