The Frontier of Data Erasure: A Survey on Machine Unlearning for Large Language Models

2025/01/01 by Youyang Qu, Ming Ding, Nan Sun +3 · 20 citations
Computer Science · Psychology · #Computer science #Data science #Erasure #Frontier #History #Machine Learning and Data Classification #Natural Language Processing Techniques #Natural language processing #Programming language #Psychology #Topic Modeling

paper · doi:10.1109/mc.2024.3405397

published in Computer 58(1), 45-57 (IEEE Computer Society)

openalex publication_date 2025/01/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/30

Abstract

This article reviews the latest in machine unlearning for large language models (LLMs), introducing methods for the targeted forgetting of information to address privacy, ethical, and legal challenges without necessitating full-model retraining. Machine unlearning is a cutting-edge solution, mitigating risk concerns of LLMs and artificial intelligence.

Citations

• Learn to Unlearn: Insights Into Machine Unlearning

Cited by

• Feature-Selective Representation Misdirection for Machine Unlearning
• Tutorial on Large Language Model-Enhanced Reinforcement Learning for Wireless Networks
• A Survey on Unlearning in Large Language Models
• Superior resilience to poisoning and amenability to unlearning in quantum machine learning
• Generative AI-Empowered Secure Communications in Space-Air-Ground Integrated Networks: A Survey and Tutorial
• Towards Reliable Forgetting: A Survey on Machine Unlearning Verification
• Bias Amplification in RAG: Poisoning Knowledge Retrieval to Steer LLMs
• UNO: Unlearning via Orthogonalization in Generative models
• Superior resilience to poisoning and amenability to unlearning in quantum machine learning. [europepmc]

Related

• The Frontier of Data Erasure: Machine Unlearning for Large Language Models
• Machine Unlearning for Traditional Models and Large Language Models: A Short Survey
• Rethinking Machine Unlearning for Large Language Models
• Does Machine Unlearning Truly Remove Knowledge?
• Unlearning in Large Language Models: We Are Not There Yet