The Lebesgue measure of ε-forging sets decays as O(ε) or ε^d for linear models and as ε^{(d-r)/2} under mild regularity assumptions, with vanishing probability of random sampling.
hub Canonical reference
A survey of machine unlearning
Canonical reference. 100% of citing Pith papers cite this work as background.
23
Pith papers citing it
67
external citations · Pith
Background
100% of classified citations
abstract
Today, computer systems hold large amounts of personal data. Yet while such an abundance of data allows breakthroughs in artificial intelligence, and especially machine learning (ML), its existence can be a threat to user privacy, and it can weaken the bonds of trust between humans and AI. Recent regulations now require that, on request, private information about a user must be removed from both computer systems and from ML models, i.e. ``the right to be forgotten''). While removing data from back-end databases should be straightforward, it is not sufficient in the AI context as ML models often `remember' the old data. Contemporary adversarial attacks on trained models have proven that we can learn whether an instance or an attribute belonged to the training data. This phenomenon calls for a new paradigm, namely machine unlearning, to make ML models forget about particular data. It turns out that recent works on machine unlearning have not been able to completely solve the problem due to the lack of common frameworks and resources. Therefore, this paper aspires to present a comprehensive examination of machine unlearning's concepts, scenarios, methods, and applications. Specifically, as a category collection of cutting-edge studies, the intention behind this article is to serve as a comprehensive resource for researchers and practitioners seeking an introduction to machine unlearning and its formulations, design criteria, removal requests, algorithms, and applications. In addition, we aim to highlight the key findings, current trends, and new research areas that have not yet featured the use of machine unlearning but could benefit greatly from it. We hope this survey serves as a valuable resource for ML researchers and those seeking to innovate privacy technologies. Our resources are publicly available at https://github.com/tamlhp/awesome-machine-unlearning.
hub tools
citation-role summary
background 6
citation-polarity summary
roles
background 6polarities
background 6representative citing papers
NPO enables stable unlearning of 50%+ training data in LLMs on TOFU by making collapse exponentially slower than gradient ascent, preserving sensible outputs where prior methods fail.
Inducing artificial uncertainty on trivial tasks allows training probes that achieve higher calibration on hard data than standard approaches while retaining performance on easy data.
TC-UMIA is a population-level attack using pre- and post-unlearning predictions to infer membership across forget, retain, and unseen sets, revealing added privacy leakage to retained data.
The Influence Eliminating Unlearning framework maximizes relearning convergence delay via weight decay and noise injection to remove the influence of a forgetting set while preserving accuracy on retained data.
Machine unlearning conflates reversing the influence of specific training examples (untraining) with removing the full underlying distribution or behavior (unlearning).
LLM unlearning is reframed as inadvertently installing backdoor triggers on forget-tokens; Random Noise Augmentation is introduced as a defense that improves robustness with theoretical guarantees.
Prompt-optimized suffixes plus synthetic fine-tuning recover ~82% of knowledge that multimodal unlearning methods claim to erase from MLLMs.
A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
Adaptive Unlearning suppresses package hallucinations in code-generating LLMs by 81% while preserving benchmark performance, using model-generated data and no human labels.
TokenUnlearn identifies critical tokens via masking and entropy signals then applies hard selection or soft weighting to unlearn only those tokens, yielding better forgetting and retained utility than sequence-level baselines on TOFU and WMDP.
REGLU guides LoRA-based unlearning via representation subspaces and orthogonal regularization to outperform prior methods on forget-retain trade-off in LLM benchmarks.
FIA uses contrastive concept saliency and temporal-spatial neuron identification to build unified masks that erase multiple target concepts while preserving general generation quality in diffusion models.
An empirical audit of one web-scraped ML training dataset reveals persistent PII after sanitization, which the authors combine with legal analysis to highlight privacy risks and advocate redefining 'publicly available' data for AI training.
A survey that organizes machine unlearning verification methods into behavioral and parametric categories and outlines open problems.
Verification of machine unlearning is fragile because model providers can use adversarial unlearning to pass checks while keeping data influence.
SalUn uses gradient-based weight saliency to achieve effective machine unlearning of data, classes, or concepts in image classification and generation, narrowing the gap to exact retraining.
CoALA is a modular cognitive architecture for language agents that organizes memory components, action spaces for internal and external interaction, and a generalized decision-making loop to support more systematic development of capable agents.
A system-first taxonomy and literature synthesis of multimodal unlearning across vision, language, video, and audio, with datasets, benchmarks, metrics, applications, and open challenges.
Inferring data distributions precisely allows distilling exact unlearning signals, yielding KL divergence bounds to the retrained model and outperforming competitors in three forgetting scenarios.
A LoRA-based residual feature alignment method for efficient machine unlearning on pre-trained models by targeting zero residuals on retained data and shifted residuals on unlearned data.
Proposes simple techniques and inverse Hessian approximation to enable certified unlearning for nonconvex DNN objectives, including nonconvergent training and sequential unlearning requests.
A survey classifying machine unlearning into centralized (exact and approximate), distributed/irregular data, verification, and privacy/security categories with technique overviews.
citing papers explorer
-
The Measure of Deception: An Analysis of Data Forging in Machine Unlearning
The Lebesgue measure of ε-forging sets decays as O(ε) or ε^d for linear models and as ε^{(d-r)/2} under mild regularity assumptions, with vanishing probability of random sampling.
-
Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning
NPO enables stable unlearning of 50%+ training data in LLMs on TOFU by making collapse exponentially slower than gradient ascent, preserving sensible outputs where prior methods fail.
-
Inducing Artificial Uncertainty in Language Models
Inducing artificial uncertainty on trivial tasks allows training probes that achieve higher calibration on hard data than standard approaches while retaining performance on easy data.
-
Revisiting Privacy Leakage in Machine Unlearning: Membership Inference Beyond the Forgotten Set
TC-UMIA is a population-level attack using pre- and post-unlearning predictions to infer membership across forget, retain, and unseen sets, revealing added privacy leakage to retained data.
-
Efficient Unlearning through Maximizing Relearning Convergence Delay
The Influence Eliminating Unlearning framework maximizes relearning convergence delay via weight decay and noise injection to remove the influence of a forgetting set while preserving accuracy on retained data.
-
Is your algorithm unlearning or untraining?
Machine unlearning conflates reversing the influence of specific training examples (untraining) with removing the full underlying distribution or behavior (unlearning).
-
Improving LLM Unlearning Robustness via Random Perturbations
LLM unlearning is reframed as inadvertently installing backdoor triggers on forget-tokens; Random Noise Augmentation is introduced as a defense that improves robustness with theoretical guarantees.
-
POPS: Recovering Unlearned Multi-Modality Knowledge in MLLMs with Prompt-Optimized Parameter Shaking
Prompt-optimized suffixes plus synthetic fine-tuning recover ~82% of knowledge that multimodal unlearning methods claim to erase from MLLMs.
-
Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces
A latent mediation framework with sparse autoencoders enables non-additive token-level influence attribution in LLMs by learning orthogonal features and back-propagating attributions.
-
LLM Ghostbusters: Surgical Hallucination Suppression via Adaptive Unlearning
Adaptive Unlearning suppresses package hallucinations in code-generating LLMs by 81% while preserving benchmark performance, using model-generated data and no human labels.
-
Unlearning What Matters: Token-Level Attribution for Precise Language Model Unlearning
TokenUnlearn identifies critical tokens via masking and entropy signals then applies hard selection or soft weighting to unlearn only those tokens, yielding better forgetting and retained utility than sequence-level baselines on TOFU and WMDP.
-
Representation-Guided Parameter-Efficient LLM Unlearning
REGLU guides LoRA-based unlearning via representation subspaces and orthogonal regularization to outperform prior methods on forget-retain trade-off in LLM benchmarks.
-
Forget-It-All: Multi-Concept Machine Unlearning via Concept-Aware Neuron Masking
FIA uses contrastive concept saliency and temporal-spatial neuron identification to build unified masks that erase multiple target concepts while preserving general generation quality in diffusion models.
-
A Common Pool of Privacy Problems: Legal and Technical Lessons from a Large-Scale Web-Scraped Machine Learning Dataset
An empirical audit of one web-scraped ML training dataset reveals persistent PII after sanitization, which the authors combine with legal analysis to highlight privacy risks and advocate redefining 'publicly available' data for AI training.
-
Towards Reliable Forgetting: A Survey on Machine Unlearning Verification
A survey that organizes machine unlearning verification methods into behavioral and parametric categories and outlines open problems.
-
Verification of Machine Unlearning is Fragile
Verification of machine unlearning is fragile because model providers can use adversarial unlearning to pass checks while keeping data influence.
-
SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation
SalUn uses gradient-based weight saliency to achieve effective machine unlearning of data, classes, or concepts in image classification and generation, narrowing the gap to exact retraining.
-
Cognitive Architectures for Language Agents
CoALA is a modular cognitive architecture for language agents that organizes memory components, action spaces for internal and external interaction, and a generalized decision-making loop to support more systematic development of capable agents.
-
Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks
A system-first taxonomy and literature synthesis of multimodal unlearning across vision, language, video, and audio, with datasets, benchmarks, metrics, applications, and open challenges.
-
Exact Unlearning from Proxies Induces Closeness Guarantees on Approximate Unlearning
Inferring data distributions precisely allows distilling exact unlearning signals, yielding KL divergence bounds to the retrained model and outperforming competitors in three forgetting scenarios.
-
Machine Unlearning on Pre-trained Models by Residual Feature Alignment Using LoRA
A LoRA-based residual feature alignment method for efficient machine unlearning on pre-trained models by targeting zero residuals on retained data and shifted residuals on unlearned data.
-
Towards Certified Unlearning for Deep Neural Networks
Proposes simple techniques and inverse Hessian approximation to enable certified unlearning for nonconvex DNN objectives, including nonconvergent training and sequential unlearning requests.
-
Machine Unlearning: A Comprehensive Survey
A survey classifying machine unlearning into centralized (exact and approximate), distributed/irregular data, verification, and privacy/security categories with technique overviews.