1 paper · 1 filter
Dustin Wright, Sarah Masud, Jared Moore +5
Large language models (LLMs) tend to generate homogenous texts, which may impact the diversity of knowledge generated across different outputs. Given their potential to replace exi…