mlsanityData to 7 Oct 2026

An Information Theoretic Evaluation Metric For Strong Unlearning

arXiv 2405.17878 ↗First posted 28 May 2024cs.LG cs.AIcited by 11

On the map: LLM safety and security → Machine unlearning

Abstract (arXiv)

Machine unlearning (MU) aims to remove the influence of specific data from trained models, addressing privacy concerns and ensuring compliance with regulations such as the ``right to be forgotten.'' Evaluating strong unlearning, where the unlearned model is indistinguishable from one retrained without the forgetting data, remains a significant challenge in deep neural networks (DNNs). Common black-box metrics, such as variants of membership inference attacks and accuracy comparisons, primarily assess model outputs but often fail to capture residual information in intermediate layers. To bridge this gap, we introduce the Information Difference Index (IDI), a novel white-box metric inspired by information theory. IDI quantifies retained information in intermediate features by measuring mutual information between those features and the labels to be forgotten, offering a more comprehensive assessment of unlearning efficacy. Our experiments demonstrate that IDI effectively measures the degree of unlearning across various datasets and architectures, providing a reliable tool for evaluating strong unlearning in DNNs.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Placed in Machine unlearning, one of 12 directions in LLM safety and security. Builds on 44 of its 61 references that are in the map. First posted on arXiv 28 May 2024.

Builds on

44 of 61 references.

Closest ancestors · 17

Same direction, newest first

  1. 2024Label-Agnostic Forgetting: A Supervision-Free Unlearning in Deep Models31 Mar
  2. 2024Inexact Unlearning Needs More Careful Evaluations to Avoid a False Sense of Privacy2 Mar
  3. 2023Layer Attack Unlearning: Fast and Accurate Machine Unlearning via Layer Level Attack and Knowledge Distillation28 Dec
  4. 2023SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation19 Oct
  5. 2023Fast Machine Unlearning Without Retraining Through Selective Synaptic Dampening15 Aug
  6. 2023Model Sparsity Can Simplify Machine Unlearning11 Apr
  7. 2023Boundary Unlearning21 Mar
  8. 2023Towards Unbounded Machine Unlearning20 Feb
  9. 2022Deep Regression Unlearning15 Oct
  10. 2022A Survey of Machine Unlearning6 Sep
  11. 2022Evaluating Machine Unlearning via Epistemic Uncertainty23 Aug
  12. 2022Can Bad Teaching Induce Forgetting? Unlearning in Deep Networks using an Incompetent Teacher17 May
5 more
  1. 2022Towards Adversarial Evaluations for Inexact Machine Unlearning17 Jan
  2. 2022Zero-Shot Machine Unlearning14 Jan
  3. 2021Fast Yet Effective Machine Unlearning17 Nov
  4. 2021Unrolling SGD: Understanding Factors Influencing Machine Unlearning27 Sep
  5. 2021Remember What You Want to Forget: Algorithms for Machine Unlearning4 Mar

Other references · 27

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  5. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  6. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  7. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  8. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,848
  9. 2020Supervised Contrastive Learningcited by 2,803
  10. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  11. 2019Machine Unlearningcited by 1,020
  12. 2021Membership Inference Attacks From First Principlescited by 839
15 more
  1. 2015Deep Learning and the Information Bottleneck Principlecited by 793
  2. 2019Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networkscited by 599
  3. 2019On Variational Bounds of Mutual Informationcited by 550
  4. 2019Certified Data Removal from Machine Learning Modelscited by 512
  5. 2019Making AI Forget You: Data Deletion in Machine Learningcited by 447
  6. 2022Last Layer Re-Training is Sufficient for Robustness to Spurious Correlationscited by 379
  7. 2020Amnesiac Machine Learningcited by 305
  8. 2020Descent-to-Delete: Gradient-Based Methods for Machine Unlearningcited by 303
  9. 2020On Hyperparameter Optimization of Machine Learning Algorithms: Theory and Practicecited by 181
  10. 2020Forgetting Outside the Box: Scrubbing Deep Networks of Information Accessible from Input-Output Observationscited by 171
  11. 2020Mixed-Privacy Forgetting in Deep Networkscited by 158
  12. 2020DeltaGrad: Rapid retraining of machine learning modelscited by 143
  13. 2020Machine Unlearning for Random Forestscited by 125
  14. 2020Coded Machine Unlearningcited by 35
  15. 2022Membership Inference Attack Using Self Influence Functionscited by 19

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 5; 2026: 6.

11 papers.

  1. 2025Are We Truly Forgetting? A Critical Re-examination of Machine Unlearning Evaluation Protocolscited by 16
  2. 2025Towards Reliable Forgetting: A Survey on Machine Unlearning Verificationcited by 8
  3. 2025DUSK: Do Not Unlearn Shared Knowledgecited by 4
  4. 2025Unlearning Comparator: A Visual Analytics System for Comparative Evaluation of Machine Unlearning Methodscited by 4
  5. 2026Auditing Language Model Unlearning via Information Decompositioncited by 3
  6. 2026An Illusion of Unlearning? Assessing Machine Unlearning Through Internal Representationscited by 3
  7. 2026Erase at the Core: Representation Unlearning for Machine Unlearningcited by 2
  8. 2025Efficient Machine Unlearning by Model Splitting and Core Sample Selectioncited by 1
  9. 2026MemLeak: Diagnosing Information Leaks in Multimodal Agent Memorycited by 1
  10. 2026Do Vision Models Truly Forget? New Findings from Representation-Level Certification of Visual Unlearning in Vertical Federated Learningcited by 0
  11. 2026Erased, but Not Gone: Output Forgetting Is Not True Forgettingcited by 0