Learning from *Sufficient* Rationales: Analysing the Relationship Between Explanation Faithfulness and Token-level Regularisation Strategies.

Jonathan Kamp, Lisa Beinborn, Antske Fokkens

Browse the full IJCNLP paper archive.