What and Whose Knowledge? Measuring Epistemic Diversity in Large Language Models
Abstract
AbstractLarge language models (LLMs) are increasingly used as primary knowledge sources, yet their epistemic diversity — defined as the diversity of real-world claims in their outputs — has never been measured. The dominant paradigm is that overall LLM diversity is low, but this is always with respect to a single point in time, with no reference baseline or consideration for variation across countries. We address this by performing the first systematic study of epistemic diversity in LLMs across time and cultural context, testing 27 LLMs on 155 topics covering 12 countries, resulting in 1.7M responses and 70M individual claims. We find that epistemic diversity has increased substantially over the past three years, a positive counter to recent diversity pessimism. However, despite progress, we find that every LLM is less diverse than a search baseline. The gap is not uniform: RAG can improve diversity, while larger models are counterintuitively less diverse than smaller ones. Moreover, LLMs systematically reflect English-language knowledge over local-language knowledge for country specific topics. Together, these results demonstrate that while progress on epistemic diversity is tangible, it is insufficient and unevenly distributed.