Forget Narrowly, Retain Broadly: Unlearning as an Asymmetric Generalization Problem

• URL: http://arxiv.org/abs/2607.09236v1
• Date: Fri, 10 Jul 2026 09:31:29 GMT
• Title: Forget Narrowly, Retain Broadly: Unlearning as an Asymmetric Generalization Problem
• Abstract summary: We develop a protocol and training corpus that captures forget-retain structure for real-world factual domains.<n>We introduce JensUn++, an unlearning algorithm that achieves the best forget-retain utility trade-off across three LLMs.
• Score: 16.147107064302435
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine unlearning in LLMs is the targeted removal of specific knowledge while preserving all other capabilities, critical for privacy and safety. Yet existing benchmarks measure it unreliably. They miss knowledge that resurfaces under paraphrased or indirect queries, a failure we call under-forgetting, and lack the semantic, syntactic, and lexical probes needed to verify that unrelated knowledge is preserved, a failure we call over-forgetting. Both failures reflect an asymmetric generalization problem. Forget evaluation must cover diverse query formulations of the same target facts, testing whether forgetting holds beyond exact training prompts. Retain evaluation must probe a far larger and implicitly defined set, namely every fact disjoint from the forget target. The retain set thus defines the effective forget set, yet current datasets provide no fine-grained annotation of this forget-retain boundary. We address this with SUITE, an evaluation protocol and training corpus that captures forget-retain structure for real-world factual domains. Methods trained on SUITE improve substantially, showing that training data is as important as algorithmic design. Building on the obtained insights, we introduce JensUn++, an unlearning algorithm that achieves the best forget-retain utility trade-off across three LLMs, in both sequential and joint unlearning settings. Code and datasets are available at https://amitpeleg.github.io/forget-narrowly-retain-broadly

Related papers

• Confess What You Know: Forget-Set Misalignment with Model Knowledge in LLM Unlearning [9.140696856930457]
  We term this gap forget-set misalignment and identify two cases.<n>In Under Unlearning, the forget set omits memorized information and leakage persists.<n>In Out-of-Knowledge Unlearning, the algorithm is driven to "forget" knowledge the model never learned, perturbing parameters and degrading utility.<n>We propose a data-blind framework that constructs model-aligned forget sets by eliciting and formalizing the model's memorized knowledge.
  arXiv  Detail & Related papers  (2026-09-01T02:47:43Z)
• PreUnlearn: Auditing Collateral Knowledge Damage Before Large Language Model Unlearning [12.982098269346773]
  We study machine unlearning for large language models from a data-centric perspective.<n>We find a consistent decay pattern: collateral damage is strongest near the forget set, weakens with semantic distance, but does not disappear at domain boundaries.<n>We formulate forget-set auditing as a pre-unlearning prediction task and analyze which data features are most predictive of downstream damage.
  arXiv  Detail & Related papers  (2026-06-16T20:32:11Z)
• How Hard Can It Be? Hardness-Aware Multi-Objective Unlearning [75.97729821186815]
  Machine unlearning aims to remove the influence of specific forget training data due to privacy, copyright or bias concerns.<n>We derive an unlearning algorithm with the overall goal of guaranteeing a specified improvement in forget quality.<n>Our hardness measure informs users when retain utility degradation is unavoidable.
  arXiv  Detail & Related papers  (2026-06-01T11:49:47Z)
• Measuring the Depth of LLM Unlearning via Activation Patching [3.590648050085134]
  Large language model (LLM) unlearning has emerged as a crucial post-hoc mechanism for privacy protection and AI safety.<n>Existing output-level metrics fail to detect when this knowledge remains recoverable from internal representations.<n>We propose the Unlearning Depth Score (UDS), a metric that quantifies the mechanistic depth of unlearning via activation patching.
  arXiv  Detail & Related papers  (2026-05-23T14:52:28Z)
• Unlearners Can Lie: Evaluating and Improving Honesty in LLM Unlearning [19.83087496179494]
  Unlearning in large language models (LLMs) aims to remove harmful training data while preserving overall utility.<n>Existing methods often hallucinate, generate abnormal token sequences, or behave inconsistently, raising safety and trust concerns.<n>We propose a formal definition of unlearning honesty, which includes preserving both utility and honesty on retained knowledge.
  arXiv  Detail & Related papers  (2026-05-09T07:50:27Z)
• SHRED: Retain-Set-Free Unlearning via Self-Distillation with Logit Demotion [39.17638540496959]
  SHRED (Self-distillation via High-surprisal-only Retain-set-free Entropy Demotion) is a retain-set-free unlearning method built on a key insight.<n>High-information tokens concentrate the model's memorized knowledge, while low-information tokens reflect general language competence.<n> SHRED is robust against relearning attacks and membership-inference attacks, and it maintains stable utility even after many sequential unlearning runs.
  arXiv  Detail & Related papers  (2026-05-08T09:25:18Z)
• LLM Unlearning on Noisy Forget Sets: A Study of Incomplete, Rewritten, and Watermarked Data [69.5099112089508]
  Large language models (LLMs) exhibit remarkable generative capabilities but raise ethical and security concerns by memorizing sensitive data.<n>This work presents the first study of unlearning under perturbed or low-fidelity forget data, referred to as noisy forget sets.<n>We find that unlearning remains surprisingly robust to perturbations, provided that core semantic signals are preserved.
  arXiv  Detail & Related papers  (2025-10-10T05:10:49Z)
• Maximally-Informative Retrieval for State Space Model Generation [59.954191072042526]
  We introduce Retrieval In-Context Optimization (RICO) to minimize model uncertainty for a particular query at test-time.<n>Unlike traditional retrieval-augmented generation (RAG), which relies on externals for document retrieval, our approach leverages direct feedback from the model.<n>We show that standard top-$k$ retrieval with model gradients can approximate our optimization procedure, and provide connections to the leave-one-out loss.
  arXiv  Detail & Related papers  (2025-06-13T18:08:54Z)
• BLUR: A Bi-Level Optimization Approach for LLM Unlearning [100.90394814817965]
  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.
  arXiv  Detail & Related papers  (2025-06-09T19:23:05Z)
• 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)
• Erasing Without Remembering: Implicit Knowledge Forgetting in Large Language Models [81.62767292169225]
  We investigate knowledge forgetting in large language models with a focus on its generalisation.<n>We propose PerMU, a novel probability perturbation-based unlearning paradigm.<n>Experiments are conducted on a diverse range of datasets, including TOFU, Harry Potter, ZsRE, WMDP, and MUSE.
  arXiv  Detail & Related papers  (2025-02-27T11:03:33Z)
• MUSE: Machine Unlearning Six-Way Evaluation for Language Models [109.76505405962783]
  Language models (LMs) are trained on vast amounts of text data, which may include private and copyrighted content. We propose MUSE, a comprehensive machine unlearning evaluation benchmark. We benchmark how effectively eight popular unlearning algorithms can unlearn Harry Potter books and news articles.
  arXiv  Detail & Related papers  (2024-07-08T23:47:29Z)
• TOFU: A Task of Fictitious Unlearning for LLMs [99.92305790945507]
  Large language models trained on massive corpora of data from the web can reproduce sensitive or private data raising both legal and ethical concerns. Unlearning, or tuning models to forget information present in their training data, provides us with a way to protect private data after training. We present TOFU, a benchmark aimed at helping deepen our understanding of unlearning.
  arXiv  Detail & Related papers  (2024-01-11T18:57:12Z)

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.