Reveal and Release: Iterative LLM Unlearning with Self-generated Data
On the map: Synthetic data and self-improvement for LLMs → Model collapse from synthetic data
also in 1 more
Abstract (arXiv)
Large language model (LLM) unlearning has demonstrated effectiveness in removing the influence of undesirable data (also known as forget data). Existing approaches typically assume full access to the forget dataset, overlooking two key challenges: (1) Forget data is often privacy-sensitive, rare, or legally regulated, making it expensive or impractical to obtain (2) The distribution of available forget data may not align with how that information is represented within the model. To address these limitations, we propose a ``Reveal-and-Release'' method to unlearn with self-generated data, where we prompt the model to reveal what it knows using optimized instructions. To fully utilize the self-generated forget data, we propose an iterative unlearning framework, where we make incremental adjustments to the model's weight space with parameter-efficient modules trained on the forget data. Experimental results demonstrate that our method balances the tradeoff between forget quality and utility preservation.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Model collapse from synthetic data, one of 13 directions in Synthetic data and self-improvement for LLMs. Builds on 21 of its 29 references that are in the map. First posted on arXiv 18 Sep 2025.
Builds on
21 of 29 references.
References · 21
- 2020Language Models are Few-Shot Learnerscited by 32,419
- 2021Training Verifiers to Solve Math Word Problemscited by 10,109
- 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
- 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
- 2021Program Synthesis with Large Language Modelscited by 3,637
- 2022Locating and Editing Factual Associations in GPTcited by 2,691
- 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,391
- 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,255
- 2024TOFU: A Task of Fictitious Unlearning for LLMscited by 533
- 2023Orthogonal Subspace Learning for Language Model Continual Learningcited by 312
- 2022The Vendi Score: A Diversity Evaluation Metric for Machine Learningcited by 311
- 2024Towards Safer Large Language Models through Machine Unlearningcited by 139
9 more
- 2021Machine Unlearning via Algorithmic Stabilitycited by 130
- 2023Separate the Wheat from the Chaff: Model Deficiency Unlearning via Parameter-Efficient Module Operationcited by 33
- 2024Unified Parameter-Efficient Unlearning for LLMscited by 25
- 2024Label-Agnostic Forgetting: A Supervision-Free Unlearning in Deep Modelscited by 18
- 2024Delta-Influence: Unlearning Poisons via Influence Functionscited by 13
- 2025Effective Skill Unlearning through Intervention and Abstentioncited by 9
- 2024Large Language Models can be Strong Self-Detoxifierscited by 7
- 2025Adversarial Mixup Unlearningcited by 7
- 2023CMD: a framework for Context-aware Model self-Detoxificationcited by 6
Built on by
Citing papers per year: 2026: 2.
2 papers.
Similar, not cited
Close in wording, no citation either way
- 2023Unlearn What You Want to Forget: Efficient Unlearning for LLMs
- 2024LLM Unlearning via Loss Adjustment with Only Forget Data
- 2025LLM Unlearning on Noisy Forget Sets: A Study of Incomplete, Rewritten, and Watermarked Data
- 2026Confess What You Know: Forget-Set Misalignment with Model Knowledge in LLM Unlearning
- 2026Unlearning in LLMs: Methods, Evaluation, and Open Challenges
- 2024Towards Robust Evaluation of Unlearning in LLMs via Data Transformations
- 2026De-attribute to Forget for LLM Unlearning
- 2024Reversing the Forget-Retain Objectives: An Efficient LLM Unlearning Framework from Logit Difference