When introducing Large Language Models (LLMs) into industrial applications, such as healthcare and education, the risk of generating harmful content becomes a significant challenge. While existing machine unlearning methods can erase specific harmful knowledge and expressions, diverse harmful content makes comprehensive removal difficult. In this study, instead of individually listing targets for forgetting, we propose Exclusive Unlearning (EU), which aims for broad harm removal by extensively forgetting everything except for the knowledge and expressions we wish to retain. We demonstrate that through Exclusive Unlearning, it is possible to obtain a model that ensures safety against a wide range of inputs, including jailbreaks, while maintaining the ability to respond to diverse instructions related to specific domains such as medicine and mathematics.
Safety Alignment via Constrained Knowledge Unlearning
Harbin Institute of Technology, Shenzhen, China(哈尔滨工业大学(深圳));University of Macau, China(澳门大学);Nankai University, China(南开大学);Zhejiang University, China(浙江大学);Peng Cheng Laboratory, China(鹏城实验室);Shaqra University, Saudi Arabia(沙克拉大学)
CRISP: Persistent Concept Unlearning via Sparse Autoencoders
Technion – Israel Institute of Technology(技术ion-以色列理工学院);Boston University(波士顿大学);University of Zagreb, FER(Zagreb大学,FER);Kempner Institute, Harvard University(哈佛大学 Kempner研究所)
Knowledge Beyond Language: Bridging the Gap in Multilingual Machine Unlearning Evaluation
GSCST, Seoul National University(首尔国立大学GSCST);AIIS, Seoul National University(首尔国立大学AIIS);Department of Artificial Intelligence, Chung-Ang University(Chung-Ang大学人工智能系);Korean Surgical Researcher Foundation, Republic of Korea(韩国外科研究员基金会)
Dissecting Fine-Tuning Unlearning in Large Language Models
International Digital Economy Academy (IDEA)(国际数字经济学院(IDEA));South China University of Technology(华南理工大学);Columbia University(哥伦比亚大学);King Abdullah University of Science and Technology(阿卜杜拉国王科技大学)
Soft Token Attacks Cannot Reliably Audit Unlearning in Large Language Models
Ludwig Maximilian University of Munich(慕尼黑路德维希-马克西米利安大学);Munich Center for Machine Learning (MCML)(慕尼黑机器学习中心);Aalto University(艾尔沃斯大学);Intel Labs(英特尔实验室)
Intrinsic Test of Unlearning Using Parametric Knowledge Traces
New York University(纽约大学);University of Toronto(多伦多大学);International Digital Economy Academy (IDEA), China(国际数字经济学院(IDEA));Bar-Ilan University(巴伊兰大学);Tel Aviv University(特拉维夫大学)
RWKU: Benchmarking Real-World Knowledge Unlearning for Large Language Models
School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院);Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所)
Consistency-Aware Editing for Entity-level Unlearning in Language Models
Shanxi University, China(山西大学);Cardiff University, UK(卡地夫大学);Singapore University of Technology and Design, Singapore(新加坡科技设计大学);ILCC, School of Informatics, University of Edinburgh, Edinburgh, UK(爱丁堡大学信息学院ILCC)
School of Computer Science, Beijing University of Posts and Telecommunications(北京邮电大学计算机科学学院);Shandong University(山东大学);New Laboratory of Pattern Recognition (NLPR), State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS), Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所)
针对 LLM 遗忘受知识表达形式制约的问题,本文提出 ORT 基准量化形式依赖偏差,并给出免训练的 ROCR 方法,通过秩一概念重定向实现更有效且自然的遗忘。
State Key Laboratory of Internet of Things for Smart City, University of Macau(物联网智能城市国家重点实验室,澳门大学);TMLR Group, Department of Computer Science, Hong Kong Baptist University(TMLR集团,香港 Baptist大学计算机科学系);Imperfect Information Learning Team, RIKEN Center for Advanced Intelligence Project(不完美信息学习团队,RIKEN高级智能项目中心);PRADA Lab, King Abdullah University of Science and Technology(PRADA实验室,国王阿卜杜勒阿齐兹大学科学与技术学院);National Institute of Informatics(国家信息研究所)
TRACER: Token ReAssignment for Concept ERasure in Generative Recommendation
Stony Brook University(石英布鲁克大学);University of Massachusetts Lowell(马萨诸塞大学洛厄尔分校);Columbia University(哥伦比亚大学);Institute of Science and Technology Austria(奥地利科学技术研究院);Sapienza University of Rome(罗马大学 sapienza);National University of Singapore(新加坡国立大学)
Unlearning Sensitive Information in Multimodal LLMs: Benchmark and Attack-Defense Evaluation
Department of Computer Science(计算机科学系);University of North Carolina at Chapel Hill(北卡罗来纳大学 Chapel Hill 分校);School of Artificial Intelligence and Data Science(人工智能与数据科学学院);University of Science and Technology of China(中国科学技术大学)