Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
41 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
41 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

SeeGULL Multilingual: a Dataset of Geo-Culturally Situated Stereotypes (2403.05696v1)

Published 8 Mar 2024 in cs.CL and cs.CV

Abstract: While generative multilingual models are rapidly being deployed, their safety and fairness evaluations are largely limited to resources collected in English. This is especially problematic for evaluations targeting inherently socio-cultural phenomena such as stereotyping, where it is important to build multi-lingual resources that reflect the stereotypes prevalent in respective language communities. However, gathering these resources, at scale, in varied languages and regions pose a significant challenge as it requires broad socio-cultural knowledge and can also be prohibitively expensive. To overcome this critical gap, we employ a recently introduced approach that couples LLM generations for scale with culturally situated validations for reliability, and build SeeGULL Multilingual, a global-scale multilingual dataset of social stereotypes, containing over 25K stereotypes, spanning 20 languages, with human annotations across 23 regions, and demonstrate its utility in identifying gaps in model evaluations. Content warning: Stereotypes shared in this paper can be offensive.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (36)
  1. Palm 2 technical report.
  2. Probing pre-trained language models for cross-cultural differences in values. In Proceedings of the First Workshop on Cross-Cultural Considerations in NLP (C3NLP), pages 114–130, Dubrovnik, Croatia. Association for Computational Linguistics.
  3. Fairmonitor: A four-stage automatic framework for detecting stereotypes and biases in large language models.
  4. Re-contextualizing fairness in nlp: The case of india. In Proceedings of the 2nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing, pages 727–740.
  5. Language models are few-shot learners. Advances in neural information processing systems, 33:1877–1901.
  6. Palm: Scaling language modeling with pathways. arXiv preprint arXiv:2204.02311.
  7. Dealing with disagreements: Looking beyond the majority vote in subjective annotations. Transactions of the Association for Computational Linguistics, 10:92–110.
  8. Building stereotype repositories with complementary approaches for scale and depth. In Proceedings of the First Workshop on Cross-Cultural Considerations in NLP (C3NLP), pages 84–90, Dubrovnik, Croatia. Association for Computational Linguistics.
  9. On measuring and mitigating biased inferences of word embeddings. Proceedings of the AAAI Conference on Artificial Intelligence, 34(05):7659–7666.
  10. Bias and fairness in large language models: A survey.
  11. Gemini Team Google. 2023. Gemini: A family of highly capable multimodal models. arXiv preprint arXiv:2312.11805.
  12. Google. 2024a. Configure safety attributes | vertex ai | google cloud.
  13. Google. 2024b. Configure safety settings for the palm api | vertex ai | google cloud.
  14. Perry Hinton. 2017. Implicit stereotypes and the predictive brain: cognition and culture in “biased” person perception. Palgrave Communications, 3(1):1–9.
  15. Dirk Hovy and Diyi Yang. 2021. The importance of modeling social factors of language: Theory and practice. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 588–602, Online. Association for Computational Linguistics.
  16. SeeGULL: A stereotype benchmark with broad geo-cultural coverage leveraging generative models. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 9851–9870, Toronto, Canada. Association for Computational Linguistics.
  17. Beyond the surface: A global-scale analysis of visual stereotypes in text-to-image generation.
  18. Otto Klineberg. 1951. The scientific study of national stereotypes. International social science bulletin, 3(3):505–514.
  19. The abc of stereotypes about groups: Agency/socioeconomic success, conservative–progressive beliefs, and communion. Journal of personality and social psychology, 110(5):675.
  20. Socially aware bias measurements for Hindi language representations. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1041–1052, Seattle, United States. Association for Computational Linguistics.
  21. Mistral AI. 2024. Endpoints | mistral ai large language models.
  22. Mistral AI. 2024. Guardrailing | mistral ai large language models.
  23. Stereoset: Measuring stereotypical bias in pretrained language models. In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 5356–5371.
  24. Socialstigmaqa: A benchmark to uncover stigma amplification in generative language models.
  25. Crows-pairs: A challenge dataset for measuring social biases in masked language models. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 1953–1967.
  26. French CrowS-pairs: Extension à une langue autre que l’anglais d’un corpus de mesure des biais sociétaux dans les modèles de langue masqués (French CrowS-pairs : Extending a challenge dataset for measuring social bias in masked language models to a language other than English). In Actes de la 29e Conférence sur le Traitement Automatique des Langues Naturelles. Volume 1 : conférence principale, pages 355–364, Avignon, France. ATALA.
  27. Gpt-4 technical report.
  28. Bbq: A hand-built bias benchmark for question answering. In Findings of the Association for Computational Linguistics: ACL 2022, pages 2086–2105.
  29. Cultural incongruencies in artificial intelligence.
  30. Stereotyping and impression formation: How categorical thinking shapes person perception. 2007) The Sage Handbook of Social Psychology: Concise Student Edition. London: Sage Publications Ltd, pages 68–92.
  31. Re-imagining algorithmic fairness in india and beyond. In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, FAccT ’21, page 315–328, New York, NY, USA. Association for Computing Machinery.
  32. Sociotechnical harms of algorithmic systems: Scoping a taxonomy for harm reduction. In Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society, pages 723–741.
  33. Mean machine translations: On gender bias in Icelandic machine translations. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 3113–3121, Marseille, France. European Language Resources Association.
  34. On evaluating and mitigating gender biases in multilingual settings. In Findings of the Association for Computational Linguistics: ACL 2023, pages 307–318, Toronto, Canada. Association for Computational Linguistics.
  35. Low-resource languages jailbreak gpt-4.
  36. Cross-cultural transfer learning for Chinese offensive language detection. In Proceedings of the First Workshop on Cross-Cultural Considerations in NLP (C3NLP), pages 8–15, Dubrovnik, Croatia. Association for Computational Linguistics.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (5)
  1. Mukul Bhutani (8 papers)
  2. Kevin Robinson (10 papers)
  3. Vinodkumar Prabhakaran (48 papers)
  4. Shachi Dave (12 papers)
  5. Sunipa Dev (28 papers)
Citations (4)
View on Semantic Scholar
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets