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 2310.20150cs.CLcs.AI
收藏

遗忘你想忘记的:面向大语言模型的高效遗忘方法

Unlearn What You Want to Forget: Efficient Unlearning for LLMs

Georgia Institute of Technology(佐治亚理工学院);Stanford University(斯坦福大学)

Jiaao Chen, Diyi Yang

AI总结:

本文提出一种基于轻量级遗忘层和选择性师生目标的高效LLM遗忘框架,并通过融合机制支持连续遗忘操作,在分类与生成任务上验证了其有效性。

摘要:

大语言模型(LLMs)通过对广泛文本数据进行预训练和记忆取得了显著进展,然而这一过程可能面临隐私问题和违反数据保护法规的风险。因此,能够轻松地从这些模型中移除与个人用户相关的数据,同时在移除后不降低其预测质量,变得越来越重要。为了解决这些问题,本文提出了一种高效的遗忘框架,该框架无需在数据移除后重新训练整个模型即可高效更新LLMs,具体做法是在Transformer中引入轻量级遗忘层,并通过选择性师生目标进行学习。此外,我们引入了一种融合机制,以有效组合学习遗忘不同数据集的多个遗忘层,从而处理一系列遗忘操作。在分类和生成任务上的实验表明,与最先进的基线方法相比,我们提出的方法具有显著的有效性。

英文摘要:

Large language models (LLMs) have achieved significant progress from pre-training on and memorizing a wide range of textual data, however, this process might suffer from privacy issues and violations of data protection regulations. As a result, the ability to easily remove data related to individual users from such models while not deteriorating their predictive quality after the removal becomes increasingly important. To address these issues, in this work, we propose an efficient unlearning framework that could efficiently update LLMs without having to retrain the whole model after data removals, by introducing lightweight unlearning layers learned with a selective teacher-student objective into the transformers. In addition, we introduce a fusion mechanism to effectively combine different unlearning layers that learns to forget different sets of data to handle a sequence of forgetting operations. Experiments on classification and generation tasks demonstrate the effectiveness of our proposed methods compared to the state-of-the-art baselines.

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。

↑