论文标题
大型预训练的语言模型是否泄漏了您的个人信息？
Are Large Pre-Trained Language Models Leaking Your Personal Information?
论文作者
论文摘要
大型预训练的语言模型是否泄漏了您的个人信息？在本文中，我们分析了预训练的语言模型（PLM）是否容易泄漏个人信息。具体来说，我们向PLMS查询电子邮件地址，其中包含电子邮件地址的上下文或包含所有者名称的提示。我们发现，由于记忆，PLM确实会泄漏个人信息。但是，由于模型在关联方面较弱，因此攻击者提取的特定个人信息的风险很低。我们希望这项工作可以帮助社区更好地了解PLM的隐私风险，并带来新的见解以确保PLM的安全。
Are Large Pre-Trained Language Models Leaking Your Personal Information? In this paper, we analyze whether Pre-Trained Language Models (PLMs) are prone to leaking personal information. Specifically, we query PLMs for email addresses with contexts of the email address or prompts containing the owner's name. We find that PLMs do leak personal information due to memorization. However, since the models are weak at association, the risk of specific personal information being extracted by attackers is low. We hope this work could help the community to better understand the privacy risk of PLMs and bring new insights to make PLMs safe.
