系统梳理大模型偏见的类型、来源与治理方法，为公平应用提供指南。
A Comprehensive Survey of Bias in LLMs: Current Landscape and Future Directions
- 按维度分类大模型偏见，涵盖内容、训练数据、评估等多个层面。
- 总结现有缓解策略并指出其局限性，推动更公平的模型设计。
- 适合研究者、开发者和政策制定者了解偏见问题与应对方向。
大型语言模型（LLMs）在自然语言处理中实现了前所未有的文本生成、翻译与理解能力，但其广泛应用暴露了模型内嵌偏见的重大隐患。本文全面综述了大模型中的偏见问题，系统梳理了偏见的类型、来源、影响及缓解策略。从多个维度对偏见进行分类，并整合当前研究成果，探讨其在真实应用中的影响。同时，批判性评估现有缓解技术的成效与不足，并提出未来研究方向，以促进大模型在公平性与公正性方面的提升。本综述为关注大模型偏见的研究人员、从业者与政策制定者提供了基础参考。
原文摘要 · Abstract (English)
Large Language Models(LLMs) have revolutionized various applications in natural language processing (NLP) by providing unprecedented text generation, translation, and comprehension capabilities. However, their widespread deployment has brought to light significant concerns regarding biases embedded within these models. This paper presents a comprehensive survey of biases in LLMs, aiming to provide an extensive review of the types, sources, impacts, and mitigation strategies related to these biases. We systematically categorize biases into several dimensions. Our survey synthesizes current research findings and discusses the implications of biases in real-world applications. Additionally, we critically assess existing bias mitigation techniques and propose future research directions to enhance fairness and equity in LLMs. This survey serves as a foundational resource for researchers, practitioners, and policymakers concerned with addressing and understanding biases in LLMs.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。