mlsanityData to 14 Oct 2026

Learn and Unlearn: Addressing Misinformation in Multilingual LLMs

arXiv 2406.13748 ↗First posted 19 Jun 2024cs.CL cs.LGcited by 12

On the map: LLM safety and security → Machine unlearning

Abstract (arXiv)

This paper investigates the propagation of harmful information in multilingual large language models (LLMs) and evaluates the efficacy of various unlearning methods. We demonstrate that fake information, regardless of the language it is in, once introduced into these models through training data, can spread across different languages, compromising the integrity and reliability of the generated content. Our findings reveal that standard unlearning techniques, which typically focus on English data, are insufficient in mitigating the spread of harmful content in multilingual contexts and could inadvertently reinforce harmful content across languages. We show that only by addressing harmful responses in both English and the original language of the harmful data can we effectively eliminate generations for all languages. This underscores the critical need for comprehensive unlearning strategies that consider the multilingual nature of modern LLMs to enhance their safety and reliability across diverse linguistic landscapes.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Placed in Machine unlearning, one of 12 directions in LLM safety and security. Builds on 22 of its 32 references that are in the map. First posted on arXiv 19 Jun 2024.

Builds on

22 of 32 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024Cross-Lingual Unlearning of Selective Knowledge in Multilingual Language Models18 Jun
  2. 2023In-Context Unlearning: Language Models as Few Shot Unlearners11 Oct
  3. 2023Who's Harry Potter? Approximate Unlearning in LLMs3 Oct

Other references · 19

  1. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  2. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,886
  3. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  4. 2022No Language Left Behind: Scaling Human-Centered Machine Translationcited by 1,312
  5. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,154
  6. 2023Bias and Fairness in Large Language Models: A Surveycited by 639
  7. 2020Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillationcited by 593
  8. 2022Memory-Based Model Editing at Scalecited by 447
  9. 2023Editing Large Language Models: Problems, Methods, and Opportunitiescited by 372
  10. 2023NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmarkcited by 348
  11. 2021Few-shot Learning with Multilingual Language Modelscited by 278
  12. 2023Knowledge Editing for Large Language Models: A Surveycited by 194
7 more
  1. 2019Cross-Lingual Ability of Multilingual BERT: An Empirical Studycited by 182
  2. 2023Not All Languages Are Created Equal in LLMs: Improving Multilingual Capability by Cross-Lingual-Thought Promptingcited by 165
  3. 2022Pretrained Language Models for Text Generation: A Surveycited by 122
  4. 2024The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contextscited by 94
  5. 2019Cross-Lingual Alignment of Contextual Word Embeddings, with Applications to Zero-shot Dependency Parsingcited by 70
  6. 2024Enhancing Multilingual Capabilities of Large Language Models through Self-Distillation from Resource-Rich Languagescited by 36
  7. 2021Improving Multilingual Translation by Representation and Gradient Regularizationcited by 31

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 9.

12 papers.

  1. 2025Soteria: Language-Specific Functional Parameter Steering for Multilingual Safety Alignmentcited by 10
  2. 2025Rethinking Cross-lingual Alignment: Balancing Transfer and Cultural Erasure in Multilingual LLMscited by 7
  3. 2025A Survey on Unlearning in Large Language Modelscited by 4
  4. 2026The Facade of Truth: Uncovering and Mitigating LLM Susceptibility to Deceptive Evidencecited by 3
  5. 2026Evaluating Cross-Lingual Unlearning in Multilingual Language Modelscited by 3
  6. 2026Layer-Targeted Multilingual Knowledge Erasure in Large Language Modelscited by 2
  7. 2026Multilingual Unlearning in LLMs: Transfer, Dynamics, and Reversibilitycited by 2
  8. 2026Beyond Cross-Lingual Transfer: Benchmarking Propagation Boundaries in Multilingual LLM Unlearningcited by 1
  9. 2026Unlearning in LLMs: Methods, Evaluation, and Open Challengescited by 0
  10. 2026QR-Erase: Efficient Subspace-Based Machine Unlearning with Layer Localizationcited by 0
  11. 2026μ2-Bench: A Multilingual Machine Unlearning Benchmarkcited by 0
  12. 2026Linguistic Loopholes in LLM Unlearning: From a 174-Language Benchmark to Coverage-Aware Unlearningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025A Comprehensive Survey of Machine Unlearning Techniques for Large Language Models
  2. 2026Multilingual Amnesia: On the Transferability of Unlearning in Multilingual LLMs
  3. 2025Uncovering the Potential Risks in Unlearning: Danger of English-only Unlearning in Multilingual LLMs
  4. 2026Unlearners Can Lie: Evaluating and Improving Honesty in LLM Unlearning
  5. 2024Split, Unlearn, Merge: Leveraging Data Attributes for More Effective Unlearning in LLMs
  6. 2024Position: LLM Unlearning Benchmarks are Weak Measures of Progress
  7. 2024Catastrophic Failure of LLM Unlearning via Quantization
  8. 2026Model Unlearning Objectives Vary for Distinct Language Functions