BriefGPT.xyz
Oct, 2023
零迁移连续提示：跨语言模型泛化任务语义
Zero-Shot Continuous Prompt Transfer: Generalizing Task Semantics Across Language Models
HTML
PDF
Zijun Wu, Yongkang Wu, Lili Mou
TL;DR
我们提出了一种零射击连续提示转移方法，通过将源提示编码为相对空间，并搜索相应的目标模型来传输，实验证实了我们方法的有效性，连续提示中的“任务语义”可以在各种语言模型中进行泛化。此外，我们发现从多个源模型中汇集“任务语义”可以进一步增强转移的泛化能力。
Abstract
prompt tuning
in
natural language processing
(NLP) has become an increasingly popular method for adapting large language models to specific tasks. However, the
→