arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.19959cs.CY
收藏

LLM中的性别偏见:更高智能的LLM并不一定能解决性别偏见与刻板印象

Gender Biases in LLMs: Higher intelligence in LLM does not necessarily solve gender bias and stereotyping

Rajesh Ranjan, Shailja Gupta, Surya Naranyan Singh

AI总结:

该研究通过基于角色的框架和中性名字方法分析1400个LLM生成角色,发现更高智能模型仍存在系统性性别偏见与职业刻板印象。

摘要:

大型语言模型(LLM)正被应用于生活的各个方面,但它们对偏见(尤其是性别刻板印象)的易感性引发了伦理担忧。本研究提出了一种新颖方法论、一个基于角色的框架以及一种中性名字方法论,以调查更高智能的LLM是否会减少此类偏见。我们分析了由两个知名LLM生成的1400个角色,结果显示,即使在具备更高智能和推理能力的LLM中,系统性偏见依然存在。o1对男性能力的评分(8.1)高于女性(7.9)和非二元性别者(7.80)。分析揭示,工程、数据和技术等领域持续存在刻板印象,男性在这些领域的存在占主导。相反,设计、艺术和营销等领域则呈现更强的女性存在感,强化了将创造力和沟通与女性联系起来的社会观念。本文提出了缓解此类性别偏见的未来方向,强调需要进一步研究以减少偏见并创建公平的AI模型。

英文摘要:

Large Language Models (LLMs) are finding applications in all aspects of life, but their susceptibility to biases, particularly gender stereotyping, raises ethical concerns. This study introduces a novel methodology, a persona-based framework, and a unisex name methodology to investigate whether higher-intelligence LLMs reduce such biases. We analyzed 1400 personas generated by two prominent LLMs, revealing that systematic biases persist even in LLMs with higher intelligence and reasoning capabilities. o1 rated males higher in competency (8.1) compared to females (7.9) and non-binary (7.80). The analysis reveals persistent stereotyping across fields like engineering, data, and technology, where the presence of males dominates. Conversely, fields like design, art, and marketing show a stronger presence of females, reinforcing societal notions that associate creativity and communication with females. This paper suggests future directions to mitigate such gender bias, reinforcing the need for further research to reduce biases and create equitable AI models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑