PRMU：面向多模态大语言模型中以人为中心的知识遗忘的无语料基准
PRMU: A Corpus-Free Benchmark for Person-Centric Knowledge Unlearning in Multimodal Large Language Models
浏览论文内容
中文总结 AI 辅助
针对现有多模态大语言模型（MLLMs）遗忘方法依赖语料的局限，提出无语料基准PRMU及基线SGPE，实验显示SGPE在目标遗忘、局部性保留等方面权衡更优。
中文摘要 AI 辅助
多模态大语言模型（MLLMs）已展现出存储和召回丰富人物相关知识的卓越能力，这引发了人们对可靠知识移除的日益关注。然而，现有的针对MLLMs的机器遗忘方法通常假设可以访问原始的遗忘语料和保留语料，但在现实的删除场景中，这些语料往往不可用。为解决这一局限，我们引入了PRMU，这是一个用于在现实的以人为中心的删除请求下评估无语料多模态遗忘的基准。PRMU聚焦于自然获取的人物相关知识，通过多样的文本和视觉探针（包括对抗性评估和细粒度局部性分析）来评估模型是否能够移除目标知识，同时保留相关知识。为推动该场景下的研究，我们进一步引入了Similarity-Gated Projection Editing（SGPE），这是一种轻量级的无语料遗忘基线，具备知识位移、受保护参数空间编辑以及感知局部性的多模态控制功能。在代表性MLLMs上开展的大量实验表明，现有的遗忘方法往往存在不利的遗忘-局部性权衡，在激进的遗忘设置下会出现显著的局部性退化，并且仍然易受多模态知识重新激活的影响。与此同时，SGPE在目标遗忘、局部性保留以及通用多模态效用之间提供了具有竞争力的权衡。我们希望PRMU能够推动未来面向现实且可扩展的多模态机器遗忘的研究。代码和数据集将在此处的URL发布。
英文摘要
Multimodal large language models (MLLMs) have demonstrated remarkable capabilities in storing and recalling rich person-related knowledge, raising increasing concerns about reliable knowledge removal. However, existing machine unlearning approaches for MLLMs typically assume access to original forget and retain corpora, which are often unavailable in realistic deletion scenarios. To address this limitation, we introduce PRMU, a benchmark for evaluating corpus-free multimodal unlearning under realistic person-centric deletion requests. PRMU focuses on naturally acquired person-related knowledge and evaluates whether models can remove target knowledge while preserving related knowledge through diverse textual and visual probes, including adversarial evaluation and fine-grained locality analysis. To facilitate research in this setting, we further introduce Similarity-Gated Projection Editing (SGPE), a lightweight corpus-free unlearning baseline with knowledge displacement, protected parameter-space editing, and locality-aware multimodal control. Extensive experiments on representative MLLMs reveal that existing unlearning methods often suffer from unfavorable forgetting-locality trade-offs, with significant locality degradation under aggressive forgetting settings, and remain vulnerable to multimodal knowledge reactivation. Meanwhile, SGPE provides a competitive trade-off between target forgetting, locality preservation, and general multimodal utility. We hope PRMU can facilitate future research toward realistic and scalable multimodal machine unlearning. Code and dataset will be released at https://github.com/2231122/PRMU.
