arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2303.13419cs.IR
收藏

用于概括和解释的模块检索

Modular Retrieval for Generalization and Interpretation

Juhao Liang, Chen Zhang, Zhengyang Tang, Jie Fu, Dawei Song, Benyou Wang

摘要:

新的检索任务不断涌现,促使新的检索模型不断发展。然而,为每个新的检索任务实例化检索模型是资源密集且耗时的,特别是对于采用大规模预训练的语言模型的检索模型。为了解决这个问题,我们转向一种新的检索范式称为模块化检索,其目的是解决新的检索任务,而不是组成多个现有的检索模块。在此基础上,我们提出了一个具有模块化提示微调的检索模型REMP。该方法通过深度提示微调,构造出服从任务属性的检索模块,并通过模块组合得到服从任务属性的检索模型。我们验证,REMP固有的模块化,不仅具有吸引力的概括性和可解释性在初步探索,但也取得了相当的性能,以国家的最先进的检索模型上的zero-shot检索基准。\footnote{Our code is available at \url{https://github.com/FreedomIntelligence/REMOP}}

英文摘要:

New retrieval tasks have always been emerging, thus urging the development of new retrieval models. However, instantiating a retrieval model for each new retrieval task is resource-intensive and time-consuming, especially for a retrieval model that employs a large-scale pre-trained language model. To address this issue, we shift to a novel retrieval paradigm called modular retrieval, which aims to solve new retrieval tasks by instead composing multiple existing retrieval modules. Built upon the paradigm, we propose a retrieval model with modular prompt tuning named REMOP. It constructs retrieval modules subject to task attributes with deep prompt tuning, and yields retrieval models subject to tasks with module composition. We validate that, REMOP inherently with modularity not only has appealing generalizability and interpretability in preliminary explorations, but also achieves comparable performance to state-of-the-art retrieval models on a zero-shot retrieval benchmark.\footnote{Our code is available at \url{https://github.com/FreedomIntelligence/REMOP}}

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑