mlsanityData to 14 Oct 2026

Iterative Multilingual Spectral Attribute Erasure

arXiv 2506.11244 ↗First posted 12 Jun 2025cs.CLcited by 1

Not on the map. Why?

Abstract (arXiv)

Multilingual representations embed words with similar meanings to share a common semantic space across languages, creating opportunities to transfer debiasing effects between languages. However, existing methods for debiasing are unable to exploit this opportunity because they operate on individual languages. We present Iterative Multilingual Spectral Attribute Erasure (IMSAE), which identifies and mitigates joint bias subspaces across multiple languages through iterative SVD-based truncation. Evaluating IMSAE across eight languages and five demographic dimensions, we demonstrate its effectiveness in both standard and zero-shot settings, where target language data is unavailable, but linguistically similar languages can be used for debiasing. Our comprehensive experiments across diverse language models (BERT, LLaMA, Mistral) show that IMSAE outperforms traditional monolingual and cross-lingual approaches while maintaining model utility.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 26 of its 41 references that are in the map. First posted on arXiv 12 Jun 2025.

Builds on

26 of 41 references.

References · 26

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2024The Llama 3 Herd of Modelscited by 14,715
  3. 2020Null It Out: Guarding Protected Attributes by Iterative Nullspace Projectioncited by 455
  4. 2019Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Settingcited by 349
  5. 2024Do Llamas Work in English? On the Latent Language of Multilingual Transformerscited by 281
  6. 2020Measuring and Reducing Gendered Correlations in Pre-trained Modelscited by 224
  7. 2020Towards Debiasing Sentence Representationscited by 171
  8. 2021Harms of Gender Exclusivity and Challenges in Non-Binary Representation in Language Technologiescited by 128
  9. 2021Sustainable Modular Debiasing of Language Modelscited by 100
  10. 2019Toward Gender-Inclusive Coreference Resolutioncited by 84
  11. 2024Fairness in Large Language Models: A Taxonomic Surveycited by 58
  12. 2024Spectral Editing of Activations for Large Language Model Alignmentcited by 56
14 more
  1. 2020Gender Bias in Multilingual Embeddings and Cross-Lingual Transfercited by 55
  2. 2020Multilingual Twitter Corpus and Baselines for Evaluating Demographic Bias in Hate Speech Recognitioncited by 43
  3. 2019Examining Gender Bias in Languages with Grammatical Gendercited by 40
  4. 2022Gold Doesn't Always Glitter: Spectral Removal of Linear and Nonlinear Guarded Attribute Informationcited by 34
  5. 2023On Evaluating and Mitigating Gender Biases in Multilingual Settingscited by 31
  6. 2022BLIND: Bias Removal With No Demographicscited by 22
  7. 2024How Do Multilingual Language Models Remember Facts?cited by 20
  8. 2023Investigating Bias in Multilingual Language Models: Cross-Lingual Transfer of Debiasing Techniquescited by 14
  9. 2019Density Matching for Bilingual Word Embeddingcited by 11
  10. 2022Analyzing Gender Representation in Multilingual Modelscited by 10
  11. 2023A Joint Matrix Factorization Analysis of Multilingual Representationscited by 5
  12. 2025Sparse Activation Editing for Reliable Instruction Following in Narrativescited by 4
  13. 2015Encoding Prior Knowledge with Eigenword Embeddingscited by 2
  14. 2024Beyond Voice Assistants: Exploring Advantages and Risks of an In-Car Social Robot in Real Driving Scenarioscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Debiasing Without Protected Attributes: Latent Concept Erasure from Textual Profilescited by 0