arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2406.14644cs.CL
收藏

揭示语言模型中的数据污染全貌:从检测到修复的综述

Unveiling the Spectrum of Data Contamination in Language Models: A Survey from Detection to Remediation

Yale University(耶鲁大学);Georgia Institute of Technology(佐治亚理工学院);MIT(麻省理工学院);Allen Institute for AI(艾伦人工智能研究所)

Chunyuan Deng, Yilun Zhao, Yuzhao Heng, Yitong Li, Jiannan Cao, Xiangru Tang, Arman Cohan

AI总结:

该综述针对大语言模型时代数据污染研究缺乏系统梳理的问题,全面梳理了数据污染的影响、检测方法的分类及优缺点,以及缓解策略,为该新兴领域的后续研究提供清晰指引。

摘要:

在大语言模型(LLM)时代,由于模型依赖大量来自互联网的训练语料,数据污染问题受到了越来越多的关注。训练语料与评估基准存在重叠的问题(即数据污染)已成为近期大量研究的焦点。这类研究旨在从不同角度识别污染、理解其影响并探索缓解策略。然而,在这一新兴领域中,缺乏能提供从基础概念到前沿洞见清晰脉络的综合性研究。因此,我们提出了一份数据污染领域的综合综述,梳理了迄今为止的关键问题、方法和研究发现,并指明了有待进一步研究与发展的方向。具体而言,我们首先考察了数据污染在不同阶段、不同形式下产生的影响。随后,我们对当前的污染检测方法进行了详细分析,对其进行分类以明确各类方法的研究重点、前提假设、优势与局限性。我们还讨论了污染缓解策略,为未来研究提供了清晰的指引。本综述简要概述了数据污染研究领域的最新进展,为后续研究工作提供了简明的参考指南。

英文摘要:

Data contamination has garnered increased attention in the era of large language models (LLMs) due to the reliance on extensive internet-derived training corpora. The issue of training corpus overlap with evaluation benchmarks--referred to as contamination--has been the focus of significant recent research. This body of work aims to identify contamination, understand its impacts, and explore mitigation strategies from diverse perspectives. However, comprehensive studies that provide a clear pathway from foundational concepts to advanced insights are lacking in this nascent field. Therefore, we present a comprehensive survey in the field of data contamination, laying out the key issues, methodologies, and findings to date, and highlighting areas in need of further research and development. In particular, we begin by examining the effects of data contamination across various stages and forms. We then provide a detailed analysis of current contamination detection methods, categorizing them to highlight their focus, assumptions, strengths, and limitations. We also discuss mitigation strategies, offering a clear guide for future research. This survey serves as a succinct overview of the most recent advancements in data contamination research, providing a straightforward guide for the benefit of future research endeavors.

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。

↑