1 paper · 1 filter
Elisha Diskind, Itamar Trainin, Uri Shaham +3
Parametric knowledge in Large Language Models is not equally accessible across languages. As a result, standard inference techniques often struggle to surface localized facts, lead…