Unfamiliar Finetuning Examples Control How Language Models Hallucinate.

Katie Kang, Eric Wallace, Claire J. Tomlin, Aviral Kumar, Sergey Levine

Browse the full NAACL paper archive.