Shadow Unlearning: A Neuro-Semantic Approach to Fidelity-Preserving Faceless Forgetting in LLMs

2026-01-07

arXiv.org

AI Summary

This work proposes the Shadow Unlearning paradigm to fulfill the “right to be forgotten” mandated by regulations such as the GDPR, enabling efficient removal of the influence of specific training data from large language models without accessing the original personally identifiable information (PII). The core method, Neuro-Semantic Projector Unlearning (NSPU), leverages neural-semantic projection to perform approximate unlearning on anonymized or synthetic forget samples, effectively balancing privacy preservation with model utility. Experimental results demonstrate that NSPU significantly outperforms existing approaches across multiple large language models, achieving stronger unlearning efficacy and enhanced privacy guarantees while preserving high-fidelity knowledge retention. Moreover, NSPU improves computational efficiency by at least an order of magnitude compared to current methods.

Technology Category

• Machine Learning: Privacy
• Natural Language Processing: (Large) Language Models
• Computer Vision: Large Vision Models

Application Category

• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Security and privacy of machine learning and AI applications

Abstract

Machine unlearning aims to selectively remove the influence of specific training samples to satisfy privacy regulations such as the GDPR's'Right to be Forgotten'. However, many existing methods require access to the data being removed, exposing it to membership inference attacks and potential misuse of Personally Identifiable Information (PII). We address this critical challenge by proposing Shadow Unlearning, a novel paradigm of approximate unlearning, that performs machine unlearning on anonymized forget data without exposing PII. We further propose a novel privacy-preserving framework, Neuro-Semantic Projector Unlearning (NSPU) to achieve Shadow unlearning. To evaluate our method, we compile Multi-domain Fictitious Unlearning (MuFU) forget set across five diverse domains and introduce an evaluation stack to quantify the trade-off between knowledge retention and unlearning effectiveness. Experimental results on various LLMs show that NSPU achieves superior unlearning performance, preserves model utility, and enhances user privacy. Additionally, the proposed approach is at least 10 times more computationally efficient than standard unlearning approaches. Our findings foster a new direction for privacy-aware machine unlearning that balances data protection and model fidelity.

Problem

• machine unlearning
• privacy-preserving
• Right to be Forgotten
• Personally Identifiable Information
• large language models

Innovation

• Shadow Unlearning
• Neuro-Semantic Projector Unlearning
• privacy-preserving machine unlearning
• fidelity-preserving forgetting
• approximate unlearning