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 2501.19202cs.CL
收藏

通过随机扰动提升大语言模型去学习鲁棒性

Improving LLM Unlearning Robustness via Random Perturbations

Japan Advanced Institute of Science and Technology(日本科学技术先进研究院);VNU University of Engineering and Technology(越南工程与技术大学);Monash University(莫纳什大学);RIKEN(日本研究机构)

Dang Huu-Tien, Hoang Thanh-Tung, Anh Bui, Minh-Phuong Nguyen, Le-Minh Nguyen, Naoya Inoue

AI总结:

本文提出将去学习过程视为后门攻击与防御问题,通过随机噪声增强方法提升去学习模型的鲁棒性,实验表明该方法有效保持去学习性能。

摘要:

本文展示当前大语言模型去学习方法本质上会降低模型鲁棒性,即使在保留查询中存在单个非对抗性遗忘标记时也会导致模型行为异常。我们提出一个新的理论框架,将去学习过程重新解释为后门攻击与防御问题:我们阐述遗忘过程无意中学习将遗忘标记(后门触发器)与目标表示(目标标签)对齐。因此,遗忘标记充当后门触发器,当在保留查询中激活时,会导致去学习模型行为紊乱,类似于成功的后门攻击。LLM去学习方法本身会污染模型,使其更容易受到遗忘标记的影响,并隐藏而非删除目标知识。为缓解遗忘过程带来的脆弱性,我们重新解释保留过程为后门防御,并提出随机噪声增强(RNA),一种轻量、模型和方法无关的改进方法,具有理论保证以提高去学习模型的鲁棒性。大量实验表明,RNA显著提高了去学习模型的鲁棒性,同时保持遗忘和保留性能。这种后门攻击-防御框架为去学习机制提供了见解,有助于未来研究方向以改进去学习鲁棒性。

英文摘要:

Here, we show that current LLM unlearning methods inherently reduce models' robustness, causing them to misbehave even when a single non-adversarial forget-token is present in the retain-query. Toward understanding underlying causes, we propose a novel theoretical framework that reframes the unlearning process as a backdoor attack and defense problem: we formulate how the forgetting process inadvertently learns to align forget-tokens (backdoor triggers) with the target-representations (target labels). As a result, forget-tokens act as backdoor triggers that, when activated in retain-queries, cause disruptions in unlearned models' behaviors, similar to successful backdoor attacks. The sense that, LLM unlearning methods themselves poison the model, make it more vulnerable to forget-tokens, and hide rather than erase target knowledge, describes their true mechanism. To mitigate the vulnerability caused by the forgetting process, we reinterpret the retaining process as a backdoor defense and propose Random Noise Augmentation (RNA), a lightweight, model and method-agnostic approach with theoretical guarantees for improving the robustness of unlearned models. Extensive experiments demonstrate that RNA significantly improves the robustness of unlearned models while preserving forget and retain performances. This backdoor attack-defense framework offers insights into the mechanism of unlearning that can shed light on future research directions for improving unlearning robustness.

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。

↑