1 paper
Olivia Macmillan-Scott, Michael Jacobs, Nils Metternich +1
Large language models (LLMs) are increasingly used to answer questions about politically contentious issues, yet evaluations typically treat a model's stance as a relatively stable…