arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.11046cs.CL
收藏

黑盒之内:检测预训练语言编码器中的数据泄露

Inside the Black Box: Detecting Data Leakage in Pre-trained Language Encoders

CISPA Helmholtz Center for Information Security(CISPA亥姆霍兹信息安全中心);Netflix Eyeline Studios(Netflix视线工作室)

Yuan Xin, Zheng Li, Ning Yu, Dingfan Chen, Mario Fritz, Michael Backes, Yang Zhang

AI总结:

本文首次系统研究预训练语言编码器通过下游模型黑盒输出导致的预训练数据成员泄露,在多种架构、任务和数据集上证实了该隐私风险的存在,并提供了相关分析以指导隐私保护实践。

摘要:

尽管预训练语言模型在自然语言处理(NLP)领域普遍存在,但由于其在大规模网络抓取数据上训练的特性,它们固有地带有隐私和版权方面的担忧。在本文中,我们开创性地系统探索了与预训练语言编码器相关的此类风险,特别关注通过从预训练语言编码器改编的下游模型暴露出的预训练数据成员泄露——这一方面在现有文献中很大程度上被忽视。我们的研究涵盖了四种预训练编码器架构、三个代表性下游任务和五个基准数据集的全面实验。有趣的是,我们的评估首次揭示了即使仅暴露下游模型的黑盒输出时也存在成员泄露,突显了远比先前假设更大的隐私风险。此外,我们提供了深入的分析和见解,以指导未来的研究人员和从业者在开发预训练语言模型时处理隐私考量。

英文摘要:

Despite being prevalent in the general field of Natural Language Processing (NLP), pre-trained language models inherently carry privacy and copyright concerns due to their nature of training on large-scale web-scraped data. In this paper, we pioneer a systematic exploration of such risks associated with pre-trained language encoders, specifically focusing on the membership leakage of pre-training data exposed through downstream models adapted from pre-trained language encoders-an aspect largely overlooked in existing literature. Our study encompasses comprehensive experiments across four types of pre-trained encoder architectures, three representative downstream tasks, and five benchmark datasets. Intriguingly, our evaluations reveal, for the first time, the existence of membership leakage even when only the black-box output of the downstream model is exposed, highlighting a privacy risk far greater than previously assumed. Alongside, we present in-depth analysis and insights toward guiding future researchers and practitioners in addressing the privacy considerations in developing pre-trained language models.

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。

↑