LLM在多语言提问下仍只认少数明星人物，存在认知偏见风险。
One world, one opinion? The superstar effect in LLM responses
- 用十种语言提问，考察LLM对全球人物的认知
- 跨语言识别高度集中，仅少数人被广泛认可
- 适合关注AI偏见与知识公平性的研究者阅读
随着大型语言模型（LLMs）改变信息传播方式，其观点可能影响广泛受众。本研究通过十种不同语言的提示，探讨了LLM对各领域代表性人物的认知。结果发现，响应多样性极低，少数人物在所有语言中均占据主导地位（即“明星效应”）。这一现象揭示了当LLM检索主观信息时，全球知识呈现被窄化的风险。
原文摘要 · Abstract (English)
As large language models (LLMs) are shaping the way information is shared and accessed online, their opinions have the potential to influence a wide audience. This study examines who the LLMs view as the most prominent figures across various fields, using prompts in ten different languages to explore the influence of linguistic diversity. Our findings reveal low diversity in responses, with a small number of figures dominating recognition across languages (also known as the "superstar effect"). These results highlight the risk of narrowing global knowledge representation when LLMs retrieve subjective information.
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。