Prompt to be Consistent is Better than Self-Consistent? Few-Shot and Zero-Shot Fact Verification with Pre-trained Language Models

Few-shot or zero-shot fact verification only relies on a few or no labeled training examples. In this paper, we propose a novel method called ProToCo, to Prompt pre-trained language models (PLMs) To be Consistent, for improving the factuality assessment capability of PLMs in the few-shot and zero-shot settings. Given a claim-evidence pair, ProToCo generates multiple variants of the claim with different relations and frames a simple consistency mechanism as constraints for making compatible predictions across these variants. We update PLMs by using parameter-efficient fine-tuning (PEFT), leading to more accurate predictions in few-shot and zero-shot fact verification tasks. Our experiments on three public verification datasets show that ProToCo significantly outperforms state-of-the-art few-shot fact verification baselines. With a small number of unlabeled instances, ProToCo also outperforms the strong zero-shot learner T0 on zero-shot verification. Compared to large PLMs using incontext learning (ICL) method, ProToCo outperforms OPT-30B and the Self-Consistencyenabled OPT-6.7B model in both few-and zeroshot settings. Original Input: Suppose Evidence. Can we infer Claim? Evidence： Coronavirus disease 2019 is a zoonotic infectious disease caused by severe acute respiratory syndrome coronavirus 2. Claim: The Coronavirus disease 2019 has a zoonotic origin. Negation Variant: Suppose Evidence. Can we infer it is false that Claim? PLM ❄Frozen❄ !Learned! Vectors Yes Yes No No ORG CON UNC NEG Yes Yes No No Maybe Maybe Yes Maybe No No No Yes Confirmation Variant: Suppose Evidence. Can we infer it is true that Claim? Uncertainty Variant: Suppose Evidence. Can we infer it is unclear that Claim?