mlsanityData to 7 Oct 2026

Multilingual Refusal Alignment for Safer Large Language Models

arXiv 2606.07535 ↗First posted 24 Apr 2026cs.CLcited by 2

On the map: LLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks

Abstract (arXiv)

As Large Language Models (LLMs) are deployed globally, ensuring their safety and alignment across multiple languages becomes paramount. However, safety behaviors often vary unpredictably between languages, posing significant challenges for consistent and ethical AI. In this work, we systematically investigate the dynamics of multilingual alignment, exploring whether single-language alignment transfers cross-lingually, how language consistency is preserved during training, and the resulting trade-offs with general knowledge capabilities. We introduce RefusEU, a novel refusal alignment dataset covering 12 European languages, including a dedicated test set for evaluating current state-of-the-art models. Our controlled Direct Preference Optimization (DPO) experiments provide two key insights: aligning models exclusively in English is insufficient to ensure cross-lingual safety, even for the same harm categories, whereas training on multilingual datasets can improve safety without degrading general performance, as measured by the Global MMLU benchmark.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in Safety alignment and jailbreak attacks, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 16 of its 26 references that are in the map. First posted on arXiv 24 Apr 2026.

Builds on

16 of 26 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmetic19 Feb

Other references · 15

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2017Proximal Policy Optimization Algorithmscited by 14,441
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  4. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  5. 2021A General Language Assistant as a Laboratory for Alignmentcited by 1,023
  6. 2024Refusal in Language Models Is Mediated by a Single Directioncited by 960
  7. 2024ORPO: Monolithic Preference Optimization without Reference Modelcited by 556
  8. 2024Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Promptscited by 189
  9. 2024Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluationcited by 181
  10. 2024SafetyPrompts: a Systematic Review of Open Datasets for Evaluating and Improving Large Language Model Safetycited by 75
  11. 2024The Multilingual Alignment Prism: Aligning Global and Local Preferences to Reduce Harmcited by 67
  12. 2025The State of Multilingual LLM Safety Research: From Measuring the Language Gap to Mitigating Itcited by 42
3 more
  1. 2024Code-Switching Red-Teaming: LLM Evaluation for Safety and Multilingual Understandingcited by 34
  2. 2025Multi-lingual Multi-turn Automated Red Teaming for LLMscited by 12
  3. 2025An Embarrassingly Simple Defense Against LLM Abliteration Attackscited by 12

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026The Illusion of Cross-Lingual Safety in Low-Resource Languagescited by 1
  2. 2026Baszta: Data-Centric Fine-Tuning of a Polish Multi-Label Safety Classifiercited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Alignment and Safety in Large Language Models: Safety Mechanisms, Training Paradigms, and Emerging Challenges
  2. 2024The Language Barrier: Dissecting Safety Challenges of LLMs in Multilingual Contexts
  3. 2026BabelSteering: Multilingual Safety Alignment via English Steering Vectors
  4. 2025The Hidden Space of Safety: Understanding Preference-Tuned LLMs in Multilingual context
  5. 2025Refusal Direction is Universal Across Safety-Aligned Languages
  6. 2026Align Once, Benefit Multilingually: Enforcing Multilingual Consistency for LLM Safety Alignment
  7. 2025MPO: Multilingual Safety Alignment via Reward Gap Optimization
  8. 2025CM-Align: Consistency-based Multilingual Alignment for Large Language Models