Machine Unlearning

Follow this topic Ask about this topic

Machine unlearning asks how to remove a person’s data, a class, or a learned concept from a model after training, often to support privacy, safety, or changing data rights. Approaches range from exact or certified deletion methods, such as data partitioning and noisy or influence-based updates, to approximate parameter edits, representation changes, and inference-time steering; some systems are designed from the outset to make deletion easier. Researchers evaluate methods against retrained models and test forgetting, retained utility, privacy attacks, recovery, and repeated deletion across language, vision, diffusion, graph, and multimodal systems. A central tension is that suppressed outputs do not necessarily mean information is erased, while stronger removal can damage related knowledge; scalable methods with reliable guarantees and low collateral harm remain a challenge.

Latest papers

351 papers56 papers in the last four weeks, up 70% on the four weeks before. 0.3% of all new papers.

Open Vocabulary Domain Unlearning
Sep 25, 2026Sumanth Udupa, Mehrtash Harandi, Yadan Luo +1VLM Unlearning Machine Unlearning
The University of Queensland · Monash University
Neuralyzing the Trace: Selective Representation-Level Unlearning with Contrastive Sparse Autoencoders
Sep 25, 2026Itai Zehavi, Fanny Jourdan, Ulrich AivodjiMachine Unlearning LLM Unlearning
Mila · ILLS · École normale supérieure Paris-Saclay +1
Machine Unlearning for Gibbs Supervised Learning Algorithms
Sep 24, 2026Yaiza Bermudez, Samir M. Perlaza, Iñaki EsnaolaMachine Unlearning
Centre Inria d’Universit´e Cˆote d’Azur, INRIA, Sophia Antipolis, France. · Laboratoire GAATI, Universit´e de la Polyn´esie franc¸aise, Fa‘a‘¯a, French Polynesia. · ECE Dept. Princeton University, Princeton, 08544 NJ, USA. +1
ForgetMimic: Motion Unlearning for Reinforcement Learning Humanoid Control
Sep 23, 2026Xukun Luan, Zhongxiang Lei, Chen Gong +3Humanoid Robot Control RL for Legged Locomotion
Beijing Institute of Technology · University of Virginia · Shandong University +1
Hidden not Deleted: How Networks Suppress Entangled Features
Sep 23, 2026Akash Samanta, Manish Pratap Singh, Debasis ChaudhuriFeature Superposition Mechanistic Interpretability
Techno India University Salt Lake, Kolkata, India · DRDO Young Scientist Laboratory - CT Chennai, India
Forget who you Forgot: Speaker Unlearning to Prevent Re-Identification in Zero-Shot Text-to-Speech
Sep 23, 2026Hyoeun Kim, Yujun Lee, Kyuhong ShimZero-Shot TTS Machine Unlearning
Sungkyunkwan University, Republic of Korea
Quantization-Robust Unlearning through the Lens of Retain-Forget Loss Landscapes Interaction
Sep 23, 2026Jialu Wang, Jianing Deng, Shuqing Luo +6Machine Unlearning LLM Unlearning
University of North Carolina at Chapel Hill · University of Pittsburgh · University of Arizona +1
Damnatio Memoriae: Adversarially and Selectively Forgetting Identities in the Embedding Space of Face Recognition Models
Sep 22, 2026Ünsal Öztürk, Vedrana Krivokuća Hahn, Sushil Bhattacharjee +1Adversarial Representation Learning Face Recognition
Idiap Research Institute, Martigny, Switzerland · UNIL, Lausanne, Switzerland
Mitigating Sequential Reappearance in Diffusion Data-Point Unlearning
Sep 21, 2026Donghyun Kim, Taehyuk Lee, Jinyeong Kim +4Memorization in Generative Models Diffusion Model Unlearning
Department of Computer Science and Engineering, Konkuk University Trustworthy Machine Learning Lab Seoul, Republic of Korea
GUARD: Natural Forgetting in Large Reasoning Models via Guided Answer-Reasoning Distillation
Sep 18, 2026Zeyu Yan, Guanghao Zhou, Minghui Qiu +2Privacy Leakage in Language Models Machine Unlearning
East China Normal University, Shanghai, China · Unaffiliated
$μ^2$ -Bench: A Multilingual Machine Unlearning Benchmark
Sep 17, 2026Kyomin Hwang, Hyeonjin Kim, Hyunho Lee +3Machine Unlearning LLM Unlearning
GSCST, Seoul National University · AIIS, Seoul National University
Not All Patches Are Equally Forgettable: Spatially Localized Domain Unlearning in Vision-Language Models
Sep 15, 2026Akanksha Singh, Vinod K. KurmiVLM Unlearning Machine Unlearning
Indian Institute of Science Education and Research Bhopal, India
Cascade: Hierarchical Recoverability Control for Large Language Model Unlearning
Sep 15, 2026Qingchen Yu, Shiying Duan, Xiaodong Li +5Privacy Leakage in Language Models Machine Unlearning
1Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing, Beihang University · School of Artificial Intelligence, Beihang University · Center for Applied Statistics, School of Statistics, Renmin University of China +1
Test-Time Unlearning via Sparse Autoencoder
Sep 14, 2026Pingzhi Li, Jinhao Duan, Vaishnav Tadiparthi +6Machine Unlearning LLM Unlearning
University of North Carolina at Chapel Hill · 2Honda Research Institute · 3Michigan State University
GRIN+: Towards Fast Yet Effective Machine Unlearning for Imbalanced Medical Data
Sep 14, 2026Minghui Huang, Junxiao WangMachine Unlearning
Guangzhou University, China
Correlation-Guided Fast Machine Unlearning via Hessian Analysis
Sep 11, 2026Ayushi Thakur, Ruchir Gupta, Amit Kumar Jaiswal +1Privacy-Preserving ML Influence Functions
Department of Computer Science and Engineering, Indian Institute of Technology (Banaras Hindu University) Varanasi, India · Halmstad University, Sweden
Do Influence-Derived Data Perturbations Enable Machine Unlearning? A Controlled Study of Three Plausible Roles
Sep 11, 2026Chenkai Wu, Chrispine Kambimbi, Qinyang Zeng +1Machine Unlearning Machine Unlearning Evaluation
Monash University · LibrAI · Tongji University +1
Certifying Concept Unlearning in Text-to-Image Diffusion Models
Sep 10, 2026Mansi, Luca Marzari, Francesco LeofanteDiffusion Model Unlearning Concept Erasure
Imperial College London · TU Wien
Published Unlearning Numbers Move Per Checkpoint, and Not Because the Removed Data Survives: An Audit of 263 Released Batch-Normalized Checkpoints
Sep 10, 2026Junlong Shen Xingyu LiBenchmark Auditing Benchmark Validity
University of Alberta Edmonton, Canada
Forgetting Only What Matters: Layer-Selective Unlearning toward Robust LLMs
Sep 9, 2026Ravi Ranjan, Olivera Kotevska, Agoritsa PolyzouMemorization in Language Models Language Model Robustness
Florida International University (FIU), Miami, FL, USA · Oak Ridge National Laboratory (ORNL), Oak Ridge, TN, USA
The BatchNorm Illusion: Diagnosing Normalization Artifacts in Machine Unlearning Evaluation
Sep 8, 2026Aaryaman Kalani, Murari Mandal, Dhruv Kumar +2Machine Unlearning Machine Unlearning Evaluation
BITS Pilani · KIIT Bhubaneswar · National University of Singapore
CUNO: Curriculum and Preference Optimization for Stable Graph Unlearning under Mass Deletion
Sep 8, 2026Chenhan Zhang, Ali Braytee, Madhushi Bandara +4Graph Unlearning Preference Optimization
University of Technology Sydney
When Retain Constraints Conflict: Mitigating Forget-Retain Interference in Tabular Data
Sep 6, 2026Zijie Liu, Jinhao Duan, Bingqi Shang +3Machine Unlearning LLM Unlearning
University of North Carolina at Chapel Hill · Michigan State University
On-the-go Forgetting without Explicit Unlearning via ERASE
Sep 5, 2026Kushal Chakrabarti, Mayank BaranwalMachine Unlearning Privacy-Preserving ML
Tata Consultancy Services Research, Mumbai · Indian Institute of Technology Bombay
Forgetting Without Restarting: Execution-State Unlearning for Stateful LLM Agents
Sep 4, 2026Chao Yao, Yangbo Wei, Zhen Huang +5LLM Unlearning LLM Agent Memory
Arizona State University, USA · Eastern Institute of Technology, Ningbo, China
OSR: Output Space Redistribution for Adaptive Label Removal in Classification Models
Sep 3, 2026Minyi Peng, Darian Gunamardi, Ivan Tjuawinata +2Machine Unlearning
Nanyang Technological University, 50 Nanyang Ave, Singapore, 639798
Entangled Representations Amplify Collateral Damage in Unlearning
Sep 2, 2026Evžen Wybitul, Tim G. J. Rudner, Christian Schroeder de WittMachine Unlearning Representation Disentanglement
University of Oxford · University of Toronto
Source-Free Class Relearning: Diagnosing Forgetting in Class Unlearning
Sep 2, 2026Zahra Dehghani, Pablo Piantanida, Mohammadhadi ShateriMachine Unlearning Image Classification
LIVIA · ILLS · ´ETS Montreal +2
Confess What You Know: Forget-Set Misalignment with Model Knowledge in LLM Unlearning
Sep 1, 2026Miso Kim, Georu Lee, Seungwon Jeong +1Privacy Leakage in Language Models Machine Unlearning
Dongguk University-Seoul
Where Identity Lives: Localized, Retain-Free Identity Unlearning in Multimodal Large Language Models
Aug 31, 2026Kangwook Ko, Jaehyuk Jang, Wonjun Lee +2MLLM Unlearning Privacy Protection in VLMs
KAIST

Sep 25, 2026J/K move · Enter open · S save

Open Vocabulary Domain Unlearning

Sumanth Udupa, Mehrtash Harandi, Yadan Luo, Mahsa Baktashmotlagh

The University of Queensland · Monash University

VLM Unlearning Machine Unlearning Model Editing

Open paper AskPDF arXiv

Vision-Language Models (VLMs) exhibit remarkable zero-shot generalization, yet they often encode unwanted or hazardous stylistic domains such as idealized textbook diagrams in medical AI or cartoon vehicles in autonomous driving. Approximate Domain Unlearning (ADU) aims to selectively erase a model's recognition of a target visual domain while preserving accuracy on the remaining domains. However, existing ADU methods operate under a flawed closed-vocabulary assumption: they evaluate unlearning solely on the specific object classes seen during the unlearning fine-tuning phase. Consequently, these methods do not unlearn the domain itself; they merely overfit to seen class-domain pairs, leaving the domain easily recognizable for unseen classes and providing a false sense of removal. We argue that true domain erasure must be class-agnostic. To address this, we formalize Open-Vocabulary Domain Unlearning (OVDU), a rigorous protocol that mandates domain forgetting must transfer to held-out classes. To solve the OVDU challenge, we propose a surgical parameter-editing framework. First, a Fisher Information mask isolates domain-sensitive weights, mathematically protecting foundational zero-shot generalization. Second, our Targeted Manifold Scattering (TMS) objective uses preference-based mining to locally scatter the forget domain's stylistic geometry. Evaluated across PACS, OfficeHome, and DomainNet, our method vastly improves open-vocabulary generalization over existing baselines. Crucially, it delivers exceptional sample efficiency, outperforming peak 8-shot baseline results with only 4 shots.