mlsanityData to 7 Oct 2026

Machine Unlearning: A Comprehensive Survey

arXiv 2405.07406 ↗First posted 13 May 2024cs.CR cs.AIcited by 33

On the map: LLM safety and security → Machine unlearning

Abstract (arXiv)

As the right to be forgotten has been legislated worldwide, many studies attempt to design unlearning mechanisms to protect users' privacy when they want to leave machine learning service platforms. Specifically, machine unlearning is to make a trained model to remove the contribution of an erased subset of the training dataset. This survey aims to systematically classify a wide range of machine unlearning and discuss their differences, connections and open problems. We categorize current unlearning methods into four scenarios: centralized unlearning, distributed and irregular data unlearning, unlearning verification, and privacy and security issues in unlearning. Since centralized unlearning is the primary domain, we use two parts to introduce: firstly, we classify centralized unlearning into exact unlearning and approximate unlearning; secondly, we offer a detailed introduction to the techniques of these methods. Besides the centralized unlearning, we notice some studies about distributed and irregular data unlearning and introduce federated unlearning and graph unlearning as the two representative directions. After introducing unlearning methods, we review studies about unlearning verification. Moreover, we consider the privacy and security issues essential in machine unlearning and organize the latest related literature. Finally, we discuss the challenges of various unlearning scenarios and address the potential research directions.

Key facts. Cited by 33 papers in the map (data through 7 Oct 2026). Placed in Machine unlearning, one of 12 directions in LLM safety and security. Builds on 95 of its 172 references that are in the map. First posted on arXiv 13 May 2024.

Builds on

95 of 172 references.

Closest ancestors · 27

Same direction, newest first

  1. 2024Offset Unlearning for Large Language Models17 Apr
  2. 2024Learn What You Want to Unlearn: Unlearning Inversion Attacks against Machine Unlearning4 Apr
  3. 2024Threats, Attacks, and Defenses in Machine Unlearning: A Survey20 Mar
  4. 2024Machine Unlearning of Pre-trained Large Language Models23 Feb
  5. 2024Towards Safer Large Language Models through Machine Unlearning15 Feb
  6. 2024Rethinking Machine Unlearning for Large Language Models13 Feb
  7. 2024Selective Forgetting: Advancing Machine Unlearning Techniques and Evaluation in Language Models8 Feb
  8. 2023ERASER: Machine Unlearning in MLaaS via an Inference Serving-Aware Approach3 Nov
  9. 2023Large Language Model Unlearning14 Oct
  10. 2023A Duty to Forget, a Right to be Assured? Exposing Vulnerabilities in Machine Unlearning Services15 Sep
  11. 2023Learning to Unlearn: Instance-wise Unlearning for Pre-trained Classifiers27 Jan
  12. 2022A Survey of Machine Unlearning6 Sep
15 more
  1. 2022Evaluating Machine Unlearning via Epistemic Uncertainty23 Aug
  2. 2022Can Bad Teaching Induce Forgetting? Unlearning in Deep Networks using an Incompetent Teacher17 May
  3. 2022Deep Unlearning via Randomized Conditionally Independent Hessians15 Apr
  4. 2022Knowledge Removal in Sampling-based Bayesian Inference24 Mar
  5. 2022Markov Chain Monte Carlo-Based Machine Unlearning: Unlearning What Needs to be Forgotten28 Feb
  6. 2022Deletion Inference, Reconstruction, and Compliance in Machine (Un)Learning7 Feb
  7. 2022Zero-Shot Machine Unlearning14 Jan
  8. 2021Fast Yet Effective Machine Unlearning17 Nov
  9. 2021On the Necessity of Auditable Algorithmic Definitions for Machine Unlearning22 Oct
  10. 2021Unrolling SGD: Understanding Factors Influencing Machine Unlearning27 Sep
  11. 2021Hard to Forget: Poisoning Attacks on Certified Machine Unlearning17 Sep
  12. 2021Machine Unlearning of Features and Labels26 Aug
  13. 2021Certifiable Machine Unlearning for Linear Models29 Jun
  14. 2021Adaptive Machine Unlearning8 Jun
  15. 2021Remember What You Want to Forget: Algorithms for Machine Unlearning4 Mar

Other references · 68

  1. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,592
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  4. 2019Fast Graph Representation Learning with PyTorch Geometriccited by 2,414
  5. 2015Character-level Convolutional Networks for Text Classificationcited by 2,293
  6. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  7. 2020Open Graph Benchmark: Datasets for Machine Learning on Graphscited by 2,148
  8. 2017Gradient Episodic Memory for Continual Learningcited by 1,871
  9. 2017Understanding Black-box Predictions via Influence Functionscited by 1,714
  10. 2016Variational Inference: A Review for Statisticianscited by 1,466
  11. 2018Experience Replay for Continual Learningcited by 1,074
  12. 2019Machine Unlearningcited by 1,020
38 more
  1. 2018Pitfalls of Graph Neural Network Evaluationcited by 994
  2. 2016Revisiting Semi-Supervised Learning with Graph Embeddingscited by 970
  3. 2021Learning to Prompt for Continual Learningcited by 938
  4. 2015Image-based Recommendations on Styles and Substitutescited by 702
  5. 2019Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networkscited by 599
  6. 2019Certified Data Removal from Machine Learning Modelscited by 512
  7. 2019Making AI Forget You: Data Deletion in Machine Learningcited by 447
  8. 2017Efficient Processing of Deep Neural Networks: A Tutorial and Surveycited by 427
  9. 2015Recent Advances in Convolutional Neural Networkscited by 360
  10. 2017Adversarial Examples: Attacks and Defenses for Deep Learningcited by 350
  11. 2020Amnesiac Machine Learningcited by 305
  12. 2020Descent-to-Delete: Gradient-Based Methods for Machine Unlearningcited by 303
  13. 2021Unlearnable Examples: Making Personal Data Unexploitablecited by 188
  14. 2020When Machine Unlearning Jeopardizes Privacycited by 181
  15. 2020Forgetting Outside the Box: Scrubbing Deep Networks of Information Accessible from Input-Output Observationscited by 171
  16. 2020Mixed-Privacy Forgetting in Deep Networkscited by 158
  17. 2020DeltaGrad: Rapid retraining of machine learning modelscited by 143
  18. 2021Graph Unlearningcited by 129
  19. 2020Machine Unlearning for Random Forestscited by 125
  20. 2019Updates-Leak: Data Set Inference and Reconstruction Attacks in Online Learningcited by 117
  21. 2020Variational Bayesian Unlearningcited by 114
  22. 2022The Right to be Forgotten in Federated Learning: An Efficient Realization with Rapid Retrainingcited by 109
  23. 2021Federated Unlearning via Class-Discriminative Pruningcited by 108
  24. 2019Analyzing Information Leakage of Updates to Natural Language Modelscited by 89
  25. 2020Machine Unlearning: Linear Filtration for Logit-based Classifierscited by 89
  26. 2019On Second-Order Group Influence Functions for Black-Box Predictionscited by 78
  27. 2022Recommendation Unlearningcited by 71
  28. 2022Federated Unlearning with Knowledge Distillationcited by 70
  29. 2022Backdoor Defense with Machine Unlearningcited by 61
  30. 2023GIF: A General Graph Unlearning Strategy via Influence Functioncited by 59
  31. 2023A Survey on Federated Unlearning: Challenges, Methods, and Future Directionscited by 58
  32. 2020Learn to Forget: Machine Unlearning via Neuron Maskingcited by 56
  33. 2022PUMA: Performance Unchanged Model Augmentation for Training Data Removalcited by 56
  34. 2024Erasing Undesirable Influence in Diffusion Modelscited by 52
  35. 2024Machine Unlearning for Image-to-Image Generative Modelscited by 47
  36. 2024UnlearnCanvas: Stylized Image Dataset for Enhanced Machine Unlearning Evaluation in Diffusion Modelscited by 47
  37. 2020Formalizing Data Deletion in the Context of the Right to be Forgottencited by 46
  38. 2022Certified Graph Unlearningcited by 46

50 most cited of 68 All 68 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 4; 2025: 17; 2026: 12.

33 papers.

Sort:

By area:

  1. 2025A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deploymentcited by 110
  2. 2025Learning to Unlearn while Retaining: Combating Gradient Conflicts in Machine Unlearningcited by 19
  3. 2025Erased or Dormant? Rethinking Concept Erasure Through Reversibilitycited by 9
  4. 2024A Survey on Recommendation Unlearning: Fundamentals, Taxonomy, Evaluation, and Open Questionscited by 7
  5. 2024Forget Vectors at Play: Universal Input Perturbations Driving Machine Unlearning in Image Classificationcited by 6
  6. 2025A Survey on Unlearnable Datacited by 5
  7. 2025The Right to be Forgotten in Pruning: Unveil Machine Unlearning on Sparse Modelscited by 5
  8. 2025Gaussian Certified Unlearning in High Dimensions: A Hypothesis Testing Approachcited by 5
  9. 2025Deep Contrastive Unlearning for Language Modelscited by 4
  10. 2026Forget Many, Forget Right: Scalable and Precise Concept Unlearning in Diffusion Modelscited by 4
  11. 2026The Unseen Threat: Residual Knowledge in Machine Unlearning under Perturbed Samplescited by 4
  12. 2024Learning to Forget using Hypernetworkscited by 3
21 more
  1. 2026Parameter-Efficient Token Embedding Editing for Clinical Class-Level Unlearningcited by 3
  2. 2026Class Unlearning via Depth-Aware Removal of Forget-Specific Directionscited by 3
  3. 2026Visual-Noise Guided In-Context Distillation for Multimodal Large Language Model Unlearningcited by 3
  4. 2025Ready2Unlearn: A Learning-Time Approach for Preparing Models with Future Unlearning Readinesscited by 2
  5. 2025Efficient Machine Unlearning by Model Splitting and Core Sample Selectioncited by 1
  6. 2025Losing is for Cherishing: Data Valuation Based on Machine Unlearning and Shapley Valuecited by 1
  7. 2025Superior resilience to poisoning and amenability to unlearning in quantum machine learningcited by 1
  8. 2025CUFG: Curriculum Unlearning Guided by the Forgetting Gradientcited by 1
  9. 2025Beyond Sharp Minima: Robust LLM Unlearning via Feedback-Guided Multi-Point Optimizationcited by 1
  10. 2025Improving Unlearning with Model Updates Probably Aligned with Gradientscited by 1
  11. 2026Statistical Roughness-Informed Machine Unlearningcited by 1
  12. 2026Designing to Forget: Deep Semi-parametric Models for Unlearningcited by 1
  13. 2024Dark Miner: Defend against undesirable generation for text-to-image diffusion modelscited by 0
  14. 2025Targeted Unlearning Using Perturbed Sign Gradient Methods With Applications On Medical Imagescited by 0
  15. 2025UNO: Unlearning via Orthogonalization in Generative modelscited by 0
  16. 2025Unlearning of Knowledge Graph Embedding via Preference Optimizationcited by 0
  17. 2026WIN-U: Woodbury-Informed Newton-Unlearning as a retain-free Machine Unlearning Frameworkcited by 0
  18. 2026Understanding U.S. Users' Security and Privacy Transparency Needs for Consumer-Facing Generative AIcited by 0
  19. 2026Not Every Subject Should Stay: Machine Unlearning for Noisy Engagement Recognitioncited by 0
  20. 2026CUNO: Curriculum and Preference Optimization for Stable Graph Unlearning under Mass Deletioncited by 0
  21. 2026Unmerge: Efficient Machine Unlearning via Task Arithmeticcited by 0