BLUR: A Bi-Level Optimization Approach for LLM Unlearning

• URL: http://arxiv.org/abs/2506.08164v1
• Date: Mon, 09 Jun 2025 19:23:05 GMT
• Title: BLUR: A Bi-Level Optimization Approach for LLM Unlearning
• Authors: Hadi Reisizadeh, Jinghan Jia, Zhiqi Bu, Bhanukiran Vinzamuri, Anil Ramakrishna, Kai-Wei Chang, Volkan Cevher, Sijia Liu, Mingyi Hong,
• Abstract summary: We argue that it is important to model the hierarchical structure of the unlearning problem.<n>We propose a novel algorithm, termed Bi-Level UnleaRning (textttBLUR), which delivers superior performance.
• Score: 105.98410883830596
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Enabling large language models (LLMs) to unlearn knowledge and capabilities acquired during training has proven vital for ensuring compliance with data regulations and promoting ethical practices in generative AI. Although there are growing interests in developing various unlearning algorithms, it remains unclear how to best formulate the unlearning problem. The most popular formulation uses a weighted sum of forget and retain loss, but it often leads to performance degradation due to the inherent trade-off between forget and retain losses. In this work, we argue that it is important to model the hierarchical structure of the unlearning problem, where the forget problem (which \textit{unlearns} certain knowledge and/or capabilities) takes priority over the retain problem (which preserves model utility). This hierarchical structure naturally leads to a bi-level optimization formulation where the lower-level objective focuses on minimizing the forget loss, while the upper-level objective aims to maintain the model's utility. Based on this new formulation, we propose a novel algorithm, termed Bi-Level UnleaRning (\texttt{BLUR}), which not only possesses strong theoretical guarantees but more importantly, delivers superior performance. In particular, our extensive experiments demonstrate that \texttt{BLUR} consistently outperforms all the state-of-the-art algorithms across various unlearning tasks, models, and metrics. Codes are available at https://github.com/OptimAI-Lab/BLURLLMUnlearning.

Related papers

• Efficient Machine Unlearning via Influence Approximation [75.31015485113993]
  Influence-based unlearning has emerged as a prominent approach to estimate the impact of individual training samples on model parameters without retraining.<n>This paper establishes a theoretical link between memorizing (incremental learning) and forgetting (unlearning)<n>We introduce the Influence Approximation Unlearning algorithm for efficient machine unlearning from the incremental perspective.
  arXiv  Detail & Related papers  (2025-07-31T05:34:27Z)
• UNO: Unlearning via Orthogonalization in Generative models [0.0]
  We show that our algorithms are able to forget data while maintaining the fidelity of the original model.<n>Using MNIST and CelebA data, we demonstrate that our algorithms achieve orders of magnitude faster unlearning times than their predecessors.
  arXiv  Detail & Related papers  (2025-06-05T07:37:02Z)
• UniErase: Unlearning Token as a Universal Erasure Primitive for Language Models [54.75551043657238]
  We introduce UniErase, a novel unlearning paradigm that employs learnable parametric suffix (unlearning token) to steer language models toward targeted forgetting behaviors.<n>UniErase achieves state-of-the-art (SOTA) performance across batch, sequential, and precise unlearning under fictitious and real-world knowledge settings.
  arXiv  Detail & Related papers  (2025-05-21T15:53:28Z)
• GUARD: Generation-time LLM Unlearning via Adaptive Restriction and Detection [36.38245533018162]
  Large Language Models (LLMs) have demonstrated strong capabilities in memorizing vast amounts of knowledge across diverse domains.<n>Existing unlearning efforts typically fine-tune the model with resources such as forget data, retain data, and a calibration model.<n>We propose Generation-time Unlearning via Adaptive Restriction and Detection (GUARD), a framework that enables dynamic unlearning during LLM generation.
  arXiv  Detail & Related papers  (2025-05-19T16:26:58Z)
• ReLearn: Unlearning via Learning for Large Language Models [64.2802606302194]
  We propose ReLearn, a data augmentation and fine-tuning pipeline for effective unlearning.<n>This framework introduces Knowledge Forgetting Rate (KFR) and Knowledge Retention Rate (KRR) to measure knowledge-level preservation.<n>Our experiments show that ReLearn successfully achieves targeted forgetting while preserving high-quality output.
  arXiv  Detail & Related papers  (2025-02-16T16:31:00Z)
• Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning [27.991291785091736]
  This work studies the problem of large language model (LLM) unlearning, aiming to remove unwanted data influences.<n>Despite the increasing demand for unlearning, a technically-grounded optimization framework is lacking.<n>We propose a simple yet effective unlearning optimization framework, called SimNPO, showing that simplicity' in removing the reliance on a reference model benefits unlearning.
  arXiv  Detail & Related papers  (2024-10-09T17:58:12Z)
• Towards Robust and Parameter-Efficient Knowledge Unlearning for LLMs [25.91643745340183]
  Large Language Models (LLMs) have demonstrated strong reasoning and memorization capabilities via pretraining on massive textual corpora.<n>This poses risk of privacy and copyright violations, highlighting the need for efficient machine unlearning methods.<n>We propose Low-rank Knowledge Unlearning (LoKU), a novel framework that enables robust and efficient unlearning for LLMs.
  arXiv  Detail & Related papers  (2024-08-13T04:18:32Z)
• Towards Robust Continual Learning with Bayesian Adaptive Moment Regularization [51.34904967046097]
  Continual learning seeks to overcome the challenge of catastrophic forgetting, where a model forgets previously learnt information. We introduce a novel prior-based method that better constrains parameter growth, reducing catastrophic forgetting. Results show that BAdam achieves state-of-the-art performance for prior-based methods on challenging single-headed class-incremental experiments.
  arXiv  Detail & Related papers  (2023-09-15T17:10:51Z)
• Adversarial Auto-Augment with Label Preservation: A Representation Learning Principle Guided Approach [95.74102207187545]
  We show that a prior-free autonomous data augmentation's objective can be derived from a representation learning principle. We then propose a practical surrogate to the objective that can be efficiently optimized and integrated seamlessly into existing methods.
  arXiv  Detail & Related papers  (2022-11-02T02:02:51Z)
• Communication-Efficient Robust Federated Learning with Noisy Labels [144.31995882209932]
  Federated learning (FL) is a promising privacy-preserving machine learning paradigm over distributed located data. We propose a learning-based reweighting approach to mitigate the effect of noisy labels in FL. Our approach has shown superior performance on several real-world datasets compared to various baselines.
  arXiv  Detail & Related papers  (2022-06-11T16:21:17Z)
• Self-Supervised Learning Aided Class-Incremental Lifelong Learning [17.151579393716958]
  We study the issue of catastrophic forgetting in class-incremental learning (Class-IL) In training procedure of Class-IL, as the model has no knowledge about following tasks, it would only extract features necessary for tasks learned so far, whose information is insufficient for joint classification. We propose to combine self-supervised learning, which can provide effective representations without requiring labels, with Class-IL to partly get around this problem.
  arXiv  Detail & Related papers  (2020-06-10T15:15:27Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.