1 paper
Giulio Pelosio, Devesh Batra, Noémie Bovey +4
Large Language Models (LLMs) can exhibit latent biases towards specific nationalities even when explicit demographic markers are not present. In this work, we introduce a novel nam…