arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2512.13352cs.LGcs.CLcs.CR
收藏

针对大型语言模型中成员推断在目标数据提取中的有效性研究

On the Effectiveness of Membership Inference in Targeted Data Extraction from Large Language Models

American University of Beirut(贝鲁特美国大学)

Ali Al Sahili, Ali Chehab, Razane Tajeddine

AI总结:

本研究通过整合多种成员推断技术评估其在大型语言模型目标数据提取中的有效性,并与传统基准测试进行比较以评估实际应用价值。

摘要:

大型语言模型(LLMs)容易记住训练数据,这带来了严重的隐私风险。最突出的担忧是训练数据提取和成员推断攻击(MIAs)。先前的研究表明,这些威胁是相互关联的:攻击者可以通过查询模型生成大量文本来从LLM中提取训练数据,然后应用MIAs来验证特定数据点是否包含在训练集中。在本研究中,我们将多种MIAs技术整合到数据提取流程中,以系统地评估其有效性。然后,我们将这些技术在整合设置中的性能与传统MIAs基准测试结果进行比较,从而评估它们在实际提取场景中的实用价值。

英文摘要:

Large Language Models (LLMs) are prone to memorizing training data, which poses serious privacy risks. Two of the most prominent concerns are training data extraction and Membership Inference Attacks (MIAs). Prior research has shown that these threats are interconnected: adversaries can extract training data from an LLM by querying the model to generate a large volume of text and subsequently applying MIAs to verify whether a particular data point was included in the training set. In this study, we integrate multiple MIA techniques into the data extraction pipeline to systematically benchmark their effectiveness. We then compare their performance in this integrated setting against results from conventional MIA benchmarks, allowing us to evaluate their practical utility in real-world extraction scenarios.

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。

↑