1 paper
Neil Band, Xuechen Li, Tengyu Ma +1
Language models (LMs) may lead their users to make suboptimal downstream decisions when they confidently hallucinate. This issue can be mitigated by having the LM verbally convey t…