Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
110 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

KorNAT: LLM Alignment Benchmark for Korean Social Values and Common Knowledge (2402.13605v6)

Published 21 Feb 2024 in cs.CL

Abstract: For LLMs to be effectively deployed in a specific country, they must possess an understanding of the nation's culture and basic knowledge. To this end, we introduce National Alignment, which measures an alignment between an LLM and a targeted country from two aspects: social value alignment and common knowledge alignment. Social value alignment evaluates how well the model understands nation-specific social values, while common knowledge alignment examines how well the model captures basic knowledge related to the nation. We constructed KorNAT, the first benchmark that measures national alignment with South Korea. For the social value dataset, we obtained ground truth labels from a large-scale survey involving 6,174 unique Korean participants. For the common knowledge dataset, we constructed samples based on Korean textbooks and GED reference materials. KorNAT contains 4K and 6K multiple-choice questions for social value and common knowledge, respectively. Our dataset creation process is meticulously designed and based on statistical sampling theory and was refined through multiple rounds of human review. The experiment results of seven LLMs reveal that only a few models met our reference score, indicating a potential for further enhancement. KorNAT has received government approval after passing an assessment conducted by a government-affiliated organization dedicated to evaluating dataset quality. Samples and detailed evaluation protocols of our dataset can be found in https://huggingface.co/datasets/jiyounglee0523/KorNAT .

Definition Search Book Streamline Icon: https://streamlinehq.com
References (46)
  1. Palm 2 technical report. arXiv preprint arXiv:2305.10403.
  2. Lora Aroyo and Chris Welty. 2013. Crowd truth: Harnessing disagreement in crowdsourcing a relation extraction gold standard. WebSci2013. ACM, 2013(2013).
  3. A general language assistant as a laboratory for alignment. arXiv preprint arXiv:2112.00861.
  4. Piqa: Reasoning about physical commonsense in natural language. In Proceedings of the AAAI conference on artificial intelligence, volume 34, pages 7432–7439.
  5. Language models are few-shot learners. Advances in neural information processing systems, 33:1877–1901.
  6. Assessing cross-cultural alignment between ChatGPT and human societies: An empirical study. In Proceedings of the First Workshop on Cross-Cultural Considerations in NLP (C3NLP), pages 53–67, Dubrovnik, Croatia. Association for Computational Linguistics.
  7. Veronika Cheplygina and Josien PW Pluim. 2018. Crowd disagreement about medical images is informative. In Intravascular Imaging and Computer Assisted Stenting and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis: 7th Joint International Workshop, CVII-STENT 2018 and Third International Workshop, LABELS 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 16, 2018, Proceedings 3, pages 105–111. Springer.
  8. Vicuna: An open-source chatbot impressing gpt-4 with 90%* chatgpt quality.
  9. Think you have solved question answering? try arc, the ai2 reasoning challenge. arXiv preprint arXiv:1803.05457.
  10. Disentangling perceptions of offensiveness: Cultural and moral correlates. arXiv preprint arXiv:2312.06861.
  11. Dealing with disagreements: Looking beyond the majority vote in subjective annotations. Transactions of the Association for Computational Linguistics, 10:92–110.
  12. Joseph L Fleiss and Jacob Cohen. 1973. The equivalence of weighted kappa and the intraclass correlation coefficient as measures of reliability. Educational and psychological measurement, 33(3):613–619.
  13. Social chemistry 101: Learning to reason about social and moral norms. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 653–670.
  14. Iason Gabriel. 2020. Artificial intelligence, values, and alignment. Minds and machines, 30(3):411–437.
  15. Calm: A multi-task benchmark for comprehensive assessment of language model bias. arXiv preprint arXiv:2308.12539.
  16. Aligning ai with shared human values. In International Conference on Learning Representations.
  17. Measuring massive multitask language understanding. Proceedings of the International Conference on Learning Representations (ICLR).
  18. Cosmos qa: Machine reading comprehension with contextual commonsense reasoning. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 2391–2401.
  19. Yufei Huang and Deyi Xiong. 2023. Cbbq: A chinese bias benchmark dataset curated with human-ai collaboration for large language models. arXiv preprint arXiv:2306.16244.
  20. Kobbq: Korean bias benchmark for question answering. arXiv preprint arXiv:2307.16778.
  21. Alignment of language agents. arXiv preprint arXiv:2103.14659.
  22. Qasc: A dataset for question answering via sentence composition. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 34, pages 8082–8090.
  23. A technical report for polyglot-ko: Open-source large-scale korean language models.
  24. Race: Large-scale reading comprehension dataset from examinations. arXiv preprint arXiv:1704.04683.
  25. Square: A large-scale dataset of sensitive questions and acceptable responses created through human-machine collaboration. In Proceedings of the 61th Annual Meeting of the Association for Computational Linguistics.
  26. Kosbi: A dataset for mitigating social bias risks towards safer large language model application. arXiv preprint arXiv:2305.17701.
  27. Kullm: Learning to construct korean instruction-following large language models. In Annual Conference on Human and Language Technology, pages 196–202. Human and Language Technology.
  28. Unqovering stereotyping biases via underspecified questions. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 3475–3489.
  29. TruthfulQA: Measuring how models mimic human falsehoods. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 3214–3252, Dublin, Ireland. Association for Computational Linguistics.
  30. What makes good in-context examples for gpt-3333? arXiv preprint arXiv:2101.06804.
  31. Can a suit of armor conduct electricity? a new dataset for open book question answering. arXiv preprint arXiv:1809.02789.
  32. Rethinking the role of demonstrations: What makes in-context learning work? arXiv preprint arXiv:2202.12837.
  33. Mikhail S Nikulin et al. 2001. Hellinger distance. Encyclopedia of mathematics, 78.
  34. Gpt-4 technical report.
  35. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35:27730–27744.
  36. Bbq: A hand-built bias benchmark for question answering. In Findings of the Association for Computational Linguistics: ACL 2022, pages 2086–2105.
  37. Elementary survey sampling. Cengage Learning.
  38. The tail wagging the dog: Dataset construction biases of social bias benchmarks. arXiv preprint arXiv:2210.10040.
  39. Irene Solaiman and Christy Dennison. 2021. Process for adapting language models to society (palms) with values-targeted datasets. Advances in Neural Information Processing Systems, 34:5861–5873.
  40. Value kaleidoscope: Engaging ai with pluralistic human values, rights, and duties. arXiv preprint arXiv:2309.00779.
  41. Stanford alpaca: An instruction-following llama model. https://github.com/tatsu-lab/stanford_alpaca.
  42. Gemini: a family of highly capable multimodal models. arXiv preprint arXiv:2312.11805.
  43. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288.
  44. Not all countries celebrate thanksgiving: On the cultural dominance in large language models. arXiv preprint arXiv:2310.12481.
  45. Hellaswag: Can a machine really finish your sentence? In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4791–4800.
  46. Don’t trust ChatGPT when your question is not in English: A study of multilingual abilities and types of LLMs. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 7915–7927, Singapore. Association for Computational Linguistics.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (7)
  1. Jiyoung Lee (42 papers)
  2. Minwoo Kim (27 papers)
  3. Seungho Kim (13 papers)
  4. Junghwan Kim (13 papers)
  5. Seunghyun Won (4 papers)
  6. Hwaran Lee (31 papers)
  7. Edward Choi (90 papers)
Citations (8)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets