提出随机掩码微调法，有效降低大模型对个人隐私信息的记忆风险。
Randomized Masked Finetuning: An Efficient Way to Mitigate Memorization of PIIs in LLMs
- 通过随机掩码方式微调模型，减少对敏感信息的记忆。
- 在Enron数据集上隐私泄露率下降超80%，性能损失仅5.73%。
- 适合关注模型隐私安全的研究者和开发者使用。
当前自然语言模型尤其是大语言模型在训练过程中容易记忆个人身份信息（PIIs），带来严重的安全与隐私风险。本文提出一种新型隐私保护微调方法——随机掩码微调（RMFT），可在最小化性能损失的前提下显著降低PII记忆。基于Enron邮件数据集的实验表明，相比基线微调，RMFT使总提取率降低80.81%，已见提取率降低80.17%；其表现优于去重方法，同时仅导致困惑度增加5.73%。我们还提出了MaxTER这一帕累托最优评估框架，用于衡量隐私与效用之间的权衡，并通过响应曲线下面积（AURC）指标验证了RMFT相较于去重方法的优越性。
原文摘要 · Abstract (English)
The current literature on memorization in Natural Language Models, especially Large Language Models (LLMs), poses severe security and privacy risks, as models tend to memorize personally identifying information (PIIs) from training data. We introduce Randomized Masked Fine-Tuning (RMFT), a novel privacy-preserving fine-tuning technique that reduces PII memorization while minimizing performance impact. Using the Enron Email Dataset, we demonstrate that RMFT achieves an 80.81% reduction in Total Extraction Rate and 80.17% reduction in Seen Extraction Rate compared to baseline fine-tuning, outperforming deduplication methods while maintaining only a 5.73% increase in perplexity. We present MaxTER, a Pareto-optimal evaluation framework for assessing privacy-utility tradeoffs, and show the performance of RMFT vs Deduplication by Area Under The Response Curve (AURC) metric.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。