Jessica Kahn

Jessica’s PhD research investigates bias and fairness in recommender systems powered by Large Language Models (LLMs). Her work examines how personal attributes, such as gender, may be represented within model prompts and internal representations, and how these can influence recommendation quality for different user groups. The research explores techniques such as probing, model steering, concept erasure, and mechanistic interpretability to identify and reduce these biases, with the aim of making LLM-based recommender systems more fair, transparent, and trustworthy.

Jessica Kahn is a scholarship recipient of the ARC Centre for Automated Decision-Making and Society (ADM+S) and is supervised by Dr. Danula Hettiachchi and Prof. Jeffrey Chan.