mlsanityData to 14 Oct 2026

Breaking mBad! Supervised Fine-tuning for Cross-Lingual Detoxification

arXiv 2505.16722 ↗First posted 22 May 2025cs.CL cs.AIcited by 3

Not on the map. Why?

Abstract (arXiv)

As large language models (LLMs) become increasingly prevalent in global applications, ensuring that they are toxicity-free across diverse linguistic contexts remains a critical challenge. We explore "Cross-lingual Detoxification", a cross-lingual paradigm that mitigates toxicity, enabling detoxification capabilities to transfer between high and low-resource languages across different script families. We analyze cross-lingual detoxification's effectiveness through 392 extensive settings to evaluate toxicity reduction in cross-distribution settings with limited data and investigate how mitigation impacts model performance on non-toxic tasks, revealing trade-offs between safety and knowledge preservation. Our code and dataset are publicly available at https://github.com/himanshubeniwal/Breaking-mBad.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 32 of its 41 references that are in the map. First posted on arXiv 22 May 2025.

Builds on

32 of 41 references.

References · 32

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  3. 2021On the Opportunities and Risks of Foundation Modelscited by 4,033
  4. 2022BLOOM: A 176B-Parameter Open-Access Multilingual Language Modelcited by 1,984
  5. 2020mT5: A massively multilingual pre-trained text-to-text transformercited by 1,818
  6. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,255
  7. 2021Ethical and social risks of harm from Language Modelscited by 966
  8. 2019Plug and Play Language Models: A Simple Approach to Controlled Text Generationcited by 799
  9. 2021DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Expertscited by 441
  10. 2020GeDi: Generative Discriminator Guided Sequence Generationcited by 378
  11. 2024A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicitycited by 185
  12. 2021Alignment of Language Agentscited by 168
20 more
  1. 2023Should ChatGPT be Biased? Challenges and Risks of Bias in Large Language Modelscited by 148
  2. 2021Bias Out-of-the-Box: An Empirical Analysis of Intersectional Occupational Biases in Popular Generative Language Modelscited by 141
  3. 2021Detoxifying Language Models Risks Marginalizing Minority Voicescited by 107
  4. 2024Detoxifying Large Language Models via Knowledge Editingcited by 87
  5. 2023Self-Detoxifying Language Models via Toxification Reversalcited by 46
  6. 2024RTP-LX: Can LLMs Evaluate Toxicity in Multilingual Scenarios?cited by 36
  7. 2024Preference Tuning For Toxicity Mitigation Generalizes Across Languagescited by 31
  8. 2022Language Anisotropic Cross-Lingual Model Editingcited by 26
  9. 2024Can LLMs Recognize Toxicity? A Structured Investigation Framework and Toxicity Metriccited by 23
  10. 2024PolygloToxicityPrompts: Multilingual Evaluation of Neural Toxic Degeneration in Large Language Modelscited by 20
  11. 2024Cross-lingual Editing in Multilingual Language Modelscited by 18
  12. 2024From One to Many: Expanding the Scope of Toxicity Mitigation in Language Modelscited by 18
  13. 2024MultiParaDetox: Extending Text Detoxification with Parallel Data to New Languagescited by 14
  14. 2024Towards Building a Robust Toxicity Predictorcited by 14
  15. 2024Multilingual and Explainable Text Detoxification with Parallel Corporacited by 12
  16. 2024Text Detoxification as Style Transfer in English and Hindicited by 11
  17. 2024DetoxLLM: A Framework for Detoxification with Explanationscited by 10
  18. 2023Added Toxicity Mitigation at Inference Time for Multimodal and Massively Multilingual Translationcited by 5
  19. 2024Cross-Lingual Transfer of Debiasing and Detoxification in Multilingual LLMs: An Extensive Investigationcited by 5
  20. 2024Cross-lingual Offensive Language Detection: A Systematic Review of Datasets, Transfer Approaches and Challengescited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Text Detoxification in isiXhosa and Yor`ub'a: A Cross-Lingual Machine Learning Approach for Low-Resource African Languagescited by 0
  2. 2026Where Does Toxicity Live? Mechanistic Localization and Targeted Suppression in Language Modelscited by 0
  3. 2026A Survey of Toxicity Detection and Mitigation Strategies for Multilingual Language Modelscited by 0