mlsanityData to 14 Oct 2026

De-attribute to Forget for LLM Unlearning

arXiv 2605.30919 ↗First posted 29 May 2026cs.LG cs.AIcited by 0

On the map: LLM safety and security → Machine unlearning

Abstract (arXiv)

The rapid development of large language models (LLMs) has raised concerns on the use of inappropriate data for training, which has led to a growing interest in LLM unlearning. Many existing LLM unlearning approaches rely on optimizing prediction loss(es), such as maximizing the loss on the forget set, but often face critical issues like over-forgetting and poor model utility. To address them, this paper novelly frames the optimization objective for LLM unlearning as one of zeroing out data attribution instead. In particular, we propose the first LLM unlearning framework based on data attribution rewards called DareU that performs reinforcement learning to update the LLM by reducing the attribution score of its generated responses (i.e., de-attributing) to the forget data owners. Empirical evaluation using an LLM classifier as an efficient approximation of attribution shows that DareU outperforms existing baselines by achieving effective unlearning while balancing forget quality and model utility well.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Machine unlearning, one of 12 directions in LLM safety and security. Builds on 42 of its 52 references that are in the map. First posted on arXiv 29 May 2026.

Builds on

42 of 52 references.

Closest ancestors · 13

Same direction, newest first

  1. 2025RULE: Reinforcement UnLEarning Achieves Forget-Retain Pareto Optimality8 Jun
  2. 2025SEPS: A Separability Measure for Robust Unlearning in LLMs20 May
  3. 2024Attribute-to-Delete: Machine Unlearning via Datamodel Matching30 Oct
  4. 2024MUSE: Machine Unlearning Six-Way Evaluation for Language Models8 Jul
  5. 2024Towards Effective Evaluations and Comparisons for LLM Unlearning Methods13 Jun
  6. 2024What makes unlearning hard and what to do about it3 Jun
  7. 2024SOUL: Unlocking the Power of Second-Order Optimization for LLM Unlearning28 Apr
  8. 2024Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning8 Apr
  9. 2024Machine Unlearning of Pre-trained Large Language Models23 Feb
  10. 2024TOFU: A Task of Fictitious Unlearning for LLMs11 Jan
  11. 2023Unlearn What You Want to Forget: Efficient Unlearning for LLMs31 Oct
  12. 2023Large Language Model Unlearning14 Oct
1 more
  1. 2023Towards Unbounded Machine Unlearning20 Feb

Other references · 29

  1. 2022Training language models to follow instructions with human feedbackcited by 15,971
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  4. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  5. 2025Qwen3 Technical Reportcited by 8,820
  6. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  7. 2020Learning to summarize from human feedbackcited by 2,588
  8. 2019Fine-Tuning Language Models from Human Preferencescited by 2,150
  9. 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
  10. 2021DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharingcited by 1,254
  11. 2024SimPO: Simple Preference Optimization with a Reference-Free Rewardcited by 963
  12. 2024TinyLlama: An Open-Source Small Language Modelcited by 618
17 more
  1. 2024LESS: Selecting Influential Data for Targeted Instruction Tuningcited by 532
  2. 2024The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearningcited by 509
  3. 2015Policy Distillationcited by 440
  4. 2023Scalable Extraction of Training Data from (Production) Language Modelscited by 434
  5. 2023Detecting Pretraining Data from Large Language Modelscited by 387
  6. 2023A Survey on Large Language Model (LLM) Security and Privacy: The Good, the Bad, and the Uglycited by 387
  7. 2019Towards Efficient Data Valuation Based on the Shapley Valuecited by 281
  8. 2024Is DPO Superior to PPO for LLM Alignment? A Comprehensive Studycited by 262
  9. 2023Can Sensitive Information Be Deleted From LLMs? Objectives for Defending Against Extraction Attackscited by 158
  10. 2020Explaining Black Box Predictions and Unveiling Data Artifacts through Influence Functionscited by 144
  11. 2022Training Data Influence Analysis and Estimation: A Surveycited by 135
  12. 2022Data Banzhaf: A Robust Data Valuation Framework for Machine Learningcited by 132
  13. 2023DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Modelscited by 123
  14. 2020FastIF: Scalable Influence Functions for Efficient Model Interpretation and Debuggingcited by 120
  15. 2024Data Shapley in One Training Runcited by 81
  16. 2024Helpful or Harmful Data? Fine-tuning-free Shapley Attribution for Explaining Language Model Predictionscited by 14
  17. 2025Where Did It Go Wrong? Attributing Undesirable LLM Behaviors via Representation Gradient Tracingcited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024LLM Unlearning via Loss Adjustment with Only Forget Data
  2. 2025Unlearning That Lasts: Utility-Preserving, Robust, and Almost Irreversible Forgetting in LLMs
  3. 2024Reversing the Forget-Retain Objectives: An Efficient LLM Unlearning Framework from Logit Difference
  4. 2024Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning
  5. 2025Reveal and Release: Iterative LLM Unlearning with Self-generated Data
  6. 2025GUARD: Guided Unlearning and Retention via Data Attribution for Large Language Models
  7. 2026Unlearning in LLMs: Methods, Evaluation, and Open Challenges
  8. 2025UIPE: Enhancing LLM Unlearning by Removing Knowledge Related to Forgetting Targets