PhD position on Responsible, Inclusive Large Language Models

12/11/2024

Large language models (LLMs) have revolutionized the field of artificial intelligence (AI). However, they exhibit the biases, stereotypes and discrimination contained in the underlying Internet data on which they have been trained. Despite their promise, their generated texts can be inappropriate to individuals who experience reading difficulties because of their lexical and syntactic complexity and length.  The project aims to create a platform to accurately measure, audit, control and improve the responsible behaviour of LLMs as well as adapt LLM’s output to different levels of reading skill, making these models more accessible.

The TALN research group at Universitat Pompeu Fabra  is looking for a talented candidate to pursue a PhD in TALN to carry our research in the area of readability, summarization and text simplification as well as to collaborate with research into factuality checking of current LLMs output. The candidate should have a Master degree in Engineering or similar and relevant and proved experience in the area of text simplification (e.g. project participation, paper publication, data creation).

The student will receive a fellowship for three years. The student will join the TALN research group, establish collaboration with other group members, and contribute with research at the international level in the form of peer-reviewed publications, data, and software development.

APPLY HERE