LLM-FACETS: A Privacy-Preserving Framework for Evaluating LLM Transparency and Accountability

Tom Lucas, Alessio Buscemi, Alfredo Capozucca +2

Assessing whether Large Language Models outputs are factually grounded, epistemically calibrated, and methodologically reproducible is a prerequisite for responsible AI deployment.…