Do Large Language Models Perform Latent Multi-Hop Reasoning without Exploiting Shortcuts?

Sohee Yang, Nora Kassner, Elena Gribovskaya, Sebastian Riedel, Mor Geva

Browse the full ACL paper archive.