Assessing and Mitigating Data Memorization Risks in Fine-Tuned Large Language Models

一、文章主要内容总结

该论文聚焦微调大型语言模型(LLMs)中的数据记忆风险,通过实证分析与方案设计,为LLM隐私保护提供了系统性解决方案,核心内容可分为以下几方面:

  1. 问题背景与研究动机:LLMs在自然语言处理领域应用广泛,但存在训练数据记忆问题,尤其在微调过程中,重复接触敏感数据会导致隐私泄露风险剧增。当前研究多关注预训练阶段的记忆问题,针对微调阶段小范围、高敏感性数据集的记忆风险研究存在缺口,且缺乏兼顾安全性与实用性的隐私保护框架,同时日益严格的AI隐私监管也推动了相关研究需求。
  2. 研究方法
    • 实验框架:选取GPT-2(1.5B参数)、Phi-3-mini(3.8B参数)、Gemma-2-2B(2B参数)三种不同规模与设计理念的LLM架构,构建含API密钥、数据库凭证、财务信息等敏感信息的合成数据集,嵌入真实对话场景以模拟实际数据模式。
    • 记忆检测协议:设计多提示变体与采样策略的增强型记忆检测协议,通过输入模型、秘密集与提示变体,统计秘密泄露数量,计算记忆率并利用bootstrap采样生成置信区间。
    • 隐私保护框架:提出四种互补的隐私保护方法,包括基于TF-IDF向量与余弦相似度的语义数据去重、在模型logits中添加拉普拉斯噪声的生成时差分隐私、基于香农熵的低熵输出过滤、结合正则表达式与机器学习分类器的模式化内容过滤。
    • 评估指标:从记忆率(成功提取秘密的百分比)、效用保留
Fine-tuning Language Models for Recipe Generation: A Comparative Analysis and Benchmark Study [PDF] [Copy] [Kimi] [REL] Authors: Anneketh Vij, Changhao Liu, Rahul Anil Nair, Theo Ho, Edward Shi, Ayan Bhowmick This research presents an exploration and study of the recipe generation task by fine-tuning various very small language models, with a focus on developing robust evaluation metrics and comparing across different language models the open-ended task of recipe generation. This study presents extensive experiments with multiple model architectures, ranging from T5-small (Raffel et al., 2023) and SmolLM-135M (Allal et al., 2024) to Phi-2 (Research, 2023),implementing both traditional NLP metrics and custom domain-specific evaluation metrics. Our novel evaluation framework incorporates recipe-specific metrics for assessing content quality and introduces an approach to allergen substitution. The results indicate that, while larger models generally perform better on standard metrics, the relationship between model size and recipe quality is more nuanced when considering domain-specific metrics. We find that SmolLM-360M and SmolLM-1.7B demonstrate comparable performance despite their size difference, while Phi-2 shows limitations in recipe generation despite its larger parameter count. Our comprehensive evaluation framework and allergen substitution system provide valuable insights for future work in recipe generation and broader NLG tasks that require domain expertise and safety considerations.
08-10
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

UnknownBody

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值