BriefGPT.xyz
Apr, 2022
零样机任务推广的提示一致性
Prompt Consistency for Zero-Shot Task Generalization
HTML
PDF
Chunting Zhou, Junxian He, Xuezhe Ma, Taylor Berg-Kirkpatrick, Graham Neubig
TL;DR
本文探讨了如何利用未标记的数据以提高自然语言处理任务的零样本性能,并通过规范提示一致性来鼓励模型的一致预测,并取得了实验结果上的进展和成果。
Abstract
One of the most impressive results of recent
nlp
history is the ability of
pre-trained language models
to solve new tasks in a zero-shot setting. To achieve this,
→