Towards Benign Memory Forgetting for Selective Multimodal Large Language Model Unlearning
面向选择性多模态大语言模型遗忘的良性记忆遗忘
机构 * Hefei University of Technology(合肥工业大学) ; Institute of Artificial Intelligence, Hefei Comprehensive National Science Center(合肥综合性国家科学中心人工智能研究院) ; University of Amsterdam(阿姆斯特丹大学)
AI总结 针对多模态大语言模型隐私泄露问题,提出S-MLLMUn Bench基准和SMFA框架,实现精确删除敏感知识同时保持模型基础能力。
Comments Accepted by ECCV 2026