Erase or Hide? Suppressing Spurious Unlearning Neurons for Robust Unlearning

• URL: http://arxiv.org/abs/2509.22263v1
• Date: Fri, 26 Sep 2025 12:27:23 GMT
• Title: Erase or Hide? Suppressing Spurious Unlearning Neurons for Robust Unlearning
• Authors: Nakyeong Yang, Dong-Kyum Kim, Jea Kwon, Minsung Kim, Kyomin Jung, Meeyoung Cha,
• Abstract summary: Large language models trained on web-scale data can memorize private or sensitive knowledge, raising significant privacy risks.<n>Some unlearning methods remain vulnerable to "relearning" during subsequent training, allowing a substantial portion of forgotten knowledge to resurface.<n>We introduce Ssiuu, a new class of unlearning methods that employs attribution-guided regularization to prevent spurious negative influence and faithfully remove target knowledge.
• Score: 31.829376135133554
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models trained on web-scale data can memorize private or sensitive knowledge, raising significant privacy risks. Although some unlearning methods mitigate these risks, they remain vulnerable to "relearning" during subsequent training, allowing a substantial portion of forgotten knowledge to resurface. In this paper, we show that widely used unlearning methods cause shallow alignment: instead of faithfully erasing target knowledge, they generate spurious unlearning neurons that amplify negative influence to hide it. To overcome this limitation, we introduce Ssiuu, a new class of unlearning methods that employs attribution-guided regularization to prevent spurious negative influence and faithfully remove target knowledge. Experimental results confirm that our method reliably erases target knowledge and outperforms strong baselines across two practical retraining scenarios: (1) adversarial injection of private data, and (2) benign attack using an instruction-following benchmark. Our findings highlight the necessity of robust and faithful unlearning methods for safe deployment of language models.

Related papers

• Understanding the Dilemma of Unlearning for Large Language Models [50.54260066313032]
  Unlearning seeks to remove specific knowledge from large language models (LLMs)<n>We propose unPact, an interpretable framework for unlearning via prompt attribution and contribution tracking.
  arXiv  Detail & Related papers  (2025-09-29T12:15:19Z)
• Step-by-Step Reasoning Attack: Revealing 'Erased' Knowledge in Large Language Models [9.719371187651591]
  Unlearning techniques suppress and leave the knowledge beneath the surface, thus making it retrievable with the right prompts.<n>We introduce a step-by-step reasoning-based black-box attack, Sleek, that systematically exposes unlearning failures.<n>Of the generated adversarial prompts, 62.5% successfully retrieved forgotten Harry Potter facts from WHP-unlearned Llama, while 50% exposed unfair suppression of retained knowledge.
  arXiv  Detail & Related papers  (2025-06-14T04:22:17Z)
• Erasing Without Remembering: Implicit Knowledge Forgetting in Large Language Models [70.78205685001168]
  We investigate knowledge forgetting in large language models with a focus on its generalisation.<n> UGBench is the first benchmark specifically designed to assess the unlearning of in-scope implicit knowledge.<n>We propose PerMU, a novel probability-based unlearning paradigm.
  arXiv  Detail & Related papers  (2025-02-27T11:03:33Z)
• Do Unlearning Methods Remove Information from Language Model Weights? [0.0]
  We show that fine-tuning on accessible facts can recover 88% of the pre-unlearning accuracy when applied to current unlearning methods for information learned during pretraining.<n>Our results also suggest that unlearning evaluations that measure unlearning robustness may overestimate robustness compared to evaluations that attempt to unlearn information learned during pretraining.
  arXiv  Detail & Related papers  (2024-10-11T14:06:58Z)
• Learn while Unlearn: An Iterative Unlearning Framework for Generative Language Models [52.40798352740857]
  We introduce the Iterative Contrastive Unlearning (ICU) framework, which consists of three core components.<n>A Knowledge Unlearning Induction module targets specific knowledge for removal using an unlearning loss.<n>A Contrastive Learning Enhancement module preserves the model's expressive capabilities against the pure unlearning goal.<n>An Iterative Unlearning Refinement module dynamically adjusts the unlearning process through ongoing evaluation and updates.
  arXiv  Detail & Related papers  (2024-07-25T07:09:35Z)
• From Theft to Bomb-Making: The Ripple Effect of Unlearning in Defending Against Jailbreak Attacks [85.84979847888157]
  Large Language Models (LLMs) are known to be vulnerable to jailbreak attacks.<n>LLMs can implicitly unlearn harmful knowledge that was not explicitly introduced during the unlearning phase.<n>We empirically validate this phenomenon, which makes unlearning-based methods able to decrease the Attack Success Rate.
  arXiv  Detail & Related papers  (2024-07-03T07:14:05Z)
• To Forget or Not? Towards Practical Knowledge Unlearning for Large Language Models [39.39428450239399]
  Large Language Models (LLMs) trained on extensive corpora inevitably retain sensitive data, such as personal privacy information and copyrighted material. Recent advancements in knowledge unlearning involve updating LLM parameters to erase specific knowledge. We introduce KnowUnDo to evaluate if the unlearning process inadvertently erases essential knowledge.
  arXiv  Detail & Related papers  (2024-07-02T03:34:16Z)
• UnUnlearning: Unlearning is not sufficient for content regulation in advanced generative AI [50.61495097098296]
  We revisit the paradigm in which unlearning is used for Large Language Models (LLMs) We introduce a concept of ununlearning, where unlearned knowledge gets reintroduced in-context. We argue that content filtering for impermissible knowledge will be required and even exact unlearning schemes are not enough for effective content regulation.
  arXiv  Detail & Related papers  (2024-06-27T10:24:35Z)
• Protecting Split Learning by Potential Energy Loss [70.81375125791979]
  We focus on the privacy leakage from the forward embeddings of split learning. We propose the potential energy loss to make the forward embeddings become more 'complicated'
  arXiv  Detail & Related papers  (2022-10-18T06:21:11Z)
• Knowledge Unlearning for Mitigating Privacy Risks in Language Models [31.322818016245087]
  We propose knowledge unlearning as an alternative method to reduce privacy risks for language models. We show that simply applying the unlikelihood training objective to target token sequences is effective at forgetting them. We show that unlearning can give a stronger empirical privacy guarantee in scenarios where the data vulnerable to extraction attacks are known a priori.
  arXiv  Detail & Related papers  (2022-10-04T10:18:11Z)
• Learning with Recoverable Forgetting [77.56338597012927]
  Learning wIth Recoverable Forgetting explicitly handles the task- or sample-specific knowledge removal and recovery. Specifically, LIRF brings in two innovative schemes, namely knowledge deposit and withdrawal. We conduct experiments on several datasets, and demonstrate that the proposed LIRF strategy yields encouraging results with gratifying generalization capability.
  arXiv  Detail & Related papers  (2022-07-17T16:42:31Z)
• Adversarial Targeted Forgetting in Regularization and Generative Based Continual Learning Models [2.8021833233819486]
  Continual (or "incremental") learning approaches are employed when additional knowledge or tasks need to be learned from subsequent batches or from streaming data. We show that an intelligent adversary can take advantage of a continual learning algorithm's capabilities of retaining existing knowledge over time. We show that the adversary can create a "false memory" about any task by inserting carefully-designed backdoor samples to the test instances of that task.
  arXiv  Detail & Related papers  (2021-02-16T18:45:01Z)

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.