1 paper
Eduardo Levy Yeyati, César Ciappa, Milagros Onofri
Recent multilingual studies show that political evaluations of large language models (LLMs) can vary across languages. This paper examines how robust this non-invariance is to ques…