RULE: Reinforcement UnLEarning Achieves Forget-Retain Pareto Optimality

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Zhang, Chenlong, Jin, Zhuoran, Yuan, Hongbang, Wei, Jiaheng, Zhou, Tong, Liu, Kang, Zhao, Jun, Chen, Yubo
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866913884960456704
author Zhang, Chenlong
Jin, Zhuoran
Yuan, Hongbang
Wei, Jiaheng
Zhou, Tong
Liu, Kang
Zhao, Jun
Chen, Yubo
author_facet Zhang, Chenlong
Jin, Zhuoran
Yuan, Hongbang
Wei, Jiaheng
Zhou, Tong
Liu, Kang
Zhao, Jun
Chen, Yubo
contents The widespread deployment of Large Language Models (LLMs) trained on massive, uncurated corpora has raised growing concerns about the inclusion of sensitive, copyrighted, or illegal content. This has led to increasing interest in LLM unlearning: the task of selectively removing specific information from a model without retraining from scratch or degrading overall utility. However, existing methods often rely on large-scale forget and retain datasets, and suffer from unnatural responses, poor generalization, or catastrophic utility loss. In this work, we propose Reinforcement UnLearning (RULE), an efficient framework that formulates unlearning as a refusal boundary optimization problem. RULE is trained with a small portion of the forget set and synthesized boundary queries, using a verifiable reward function that encourages safe refusal on forget--related queries while preserving helpful responses on permissible inputs. We provide both theoretical and empirical evidence demonstrating the effectiveness of RULE in achieving targeted unlearning without compromising model utility. Experimental results show that, with only $12%$ forget set and $8%$ synthesized boundary data, RULE outperforms existing baselines by up to $17.5%$ forget quality and $16.3%$ naturalness response while maintaining general utility, achieving forget--retain Pareto optimality. Remarkably, we further observe that RULE improves the naturalness of model outputs, enhances training efficiency, and exhibits strong generalization ability, generalizing refusal behavior to semantically related but unseen queries.
format Preprint
id arxiv_https___arxiv_org_abs_2506_07171
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RULE: Reinforcement UnLEarning Achieves Forget-Retain Pareto Optimality
Zhang, Chenlong
Jin, Zhuoran
Yuan, Hongbang
Wei, Jiaheng
Zhou, Tong
Liu, Kang
Zhao, Jun
Chen, Yubo
Computation and Language
Machine Learning
The widespread deployment of Large Language Models (LLMs) trained on massive, uncurated corpora has raised growing concerns about the inclusion of sensitive, copyrighted, or illegal content. This has led to increasing interest in LLM unlearning: the task of selectively removing specific information from a model without retraining from scratch or degrading overall utility. However, existing methods often rely on large-scale forget and retain datasets, and suffer from unnatural responses, poor generalization, or catastrophic utility loss. In this work, we propose Reinforcement UnLearning (RULE), an efficient framework that formulates unlearning as a refusal boundary optimization problem. RULE is trained with a small portion of the forget set and synthesized boundary queries, using a verifiable reward function that encourages safe refusal on forget--related queries while preserving helpful responses on permissible inputs. We provide both theoretical and empirical evidence demonstrating the effectiveness of RULE in achieving targeted unlearning without compromising model utility. Experimental results show that, with only $12%$ forget set and $8%$ synthesized boundary data, RULE outperforms existing baselines by up to $17.5%$ forget quality and $16.3%$ naturalness response while maintaining general utility, achieving forget--retain Pareto optimality. Remarkably, we further observe that RULE improves the naturalness of model outputs, enhances training efficiency, and exhibits strong generalization ability, generalizing refusal behavior to semantically related but unseen queries.
title RULE: Reinforcement UnLEarning Achieves Forget-Retain Pareto Optimality
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2506.07171