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 2404.16587cs.CLcs.AI
收藏

理解大型语言模型引发的嵌入隐私风险

Understanding Privacy Risks of Embeddings Induced by Large Language Models

Zhihao Zhu, Ninglu Shao, Defu Lian, Chenwang Wu, Zheng Liu, Yi Yang, Enhong Chen

AI总结:

本文研究了大型语言模型(LLMs)从文本嵌入中重建原始知识和预测实体属性的能力,发现其准确率显著高于传统预训练模型,凸显了LLMs在检索增强生成中引发用户隐私风险的负面后果,并提出了初步缓解策略。

摘要:

大型语言模型(LLMs)展现出通用人工智能的早期迹象,但受幻觉问题困扰。缓解幻觉的一种可行方案是将外部知识存储为嵌入,辅助LLMs进行检索增强生成。然而,该方案存在隐私风险,近期研究实验表明,预训练语言模型可从文本嵌入中部分重建原始文本。LLMs相较于传统预训练模型的显著优势可能加剧这些担忧。为此,我们研究了采用LLMs时,从这些嵌入中重建原始知识和预测实体属性的有效性。实证结果表明,无论文本是分布内还是分布外,LLMs均显著提高了两项评估任务的准确率。这凸显了LLMs危及用户隐私的更高潜在风险,强调了其广泛使用的负面后果。我们进一步讨论了缓解该风险的初步策略。

英文摘要:

Large language models (LLMs) show early signs of artificial general intelligence but struggle with hallucinations. One promising solution to mitigate these hallucinations is to store external knowledge as embeddings, aiding LLMs in retrieval-augmented generation. However, such a solution risks compromising privacy, as recent studies experimentally showed that the original text can be partially reconstructed from text embeddings by pre-trained language models. The significant advantage of LLMs over traditional pre-trained models may exacerbate these concerns. To this end, we investigate the effectiveness of reconstructing original knowledge and predicting entity attributes from these embeddings when LLMs are employed. Empirical findings indicate that LLMs significantly improve the accuracy of two evaluated tasks over those from pre-trained models, regardless of whether the texts are in-distribution or out-of-distribution. This underscores a heightened potential for LLMs to jeopardize user privacy, highlighting the negative consequences of their widespread use. We further discuss preliminary strategies to mitigate this risk.

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。

↑