SOUL：解锁二阶优化在LLM遗忘中的力量
SOUL: Unlocking the Power of Second-Order Optimization for LLM Unlearning
AI总结：
本文首次探讨了优化器选择在LLM遗忘中的重要性，提出基于二阶优化的遗忘框架SOUL，将静态影响遗忘扩展为动态迭代过程，并在多种任务和模型上显著优于传统一阶方法。
摘要：
大型语言模型（LLMs）凸显了有效遗忘机制的必要性，以遵守数据法规和合乎伦理的AI实践。LLM遗忘旨在消除不期望的数据影响及相关的模型能力，同时不损害遗忘范围之外的实用性。尽管对研究LLM遗忘的兴趣日益浓厚，但优化器选择对LLM遗忘的影响仍未被探索。在这项工作中，我们首次阐明了优化器选择在LLM遗忘中的重要性，建立了二阶优化与影响遗忘（一种使用影响函数更新模型以消除数据影响的经典方法）之间的明确联系。这一见解促使我们开发了一个基于二阶优化的LLM遗忘框架，称为Second-Order UnLearning（SOUL），它将使用影响遗忘的静态、一次性模型更新扩展为动态、迭代的遗忘过程。我们广泛的实验表明，SOUL在各种遗忘任务、模型和指标上始终优于传统的一阶方法，表明二阶优化为LLM遗忘提供了有效且广泛适用的解决方案。代码可在 https://github.com/OPTML-Group/SOUL 获取。
英文摘要：
Large Language Models (LLMs) have highlighted the necessity of effective unlearning mechanisms to comply with data regulations and ethical AI practices. LLM unlearning aims at removing undesired data influences and associated model capabilities without compromising utility beyond the scope of unlearning. While interest in studying LLM unlearning is growing, the impact of the optimizer choice for LLM unlearning remains unexplored. In this work, we shed light on the significance of optimizer selection in LLM unlearning for the first time, establishing a clear connection between second-order optimization and influence unlearning (a classical approach using influence functions to update the model for data influence removal). This insight propels us to develop a second-order optimization-based LLM unlearning framework, termed Second-Order UnLearning (SOUL), which extends the static, one-shot model update using influence unlearning to a dynamic, iterative unlearning process. Our extensive experiments show that SOUL consistently outperforms conventional first-order methods across various unlearning tasks, models, and metrics, indicating that second-order optimization offers an effective and broadly applicable solution for LLM unlearning. Codes are available at https://github.com/OPTML-Group/SOUL.
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。
