Towards Understanding Large-Scale Discourse Structures in Pre-Trained and Fine-Tuned Language Models.

Patrick Huber, Giuseppe Carenini

Browse the full NAACL paper archive.