大语言模型与全球人类观点的对齐研究
On the Alignment of Large Language Models with Global Human Opinion
AI总结：
本研究基于世界价值观调查构建评估框架，首次从全球、语言和时间维度全面探究LLM的观点对齐问题，发现调整提示词语言可更有效引导LLM对齐对应国家观点。
摘要：
当前的大语言模型（LLM）已具备支持多语言场景的能力，用户可使用母语与LLM交互。当LLM回应用户提出的主观性问题时，人们期望其能与特定人口群体或特定历史时期的观点保持对齐，而这种对齐会受到用户与模型交互所用语言的影响。现有研究主要聚焦于LLM在美国或少数国家人口群体中所体现的观点，缺乏覆盖全球各国的样本，也缺少对不同历史时期人类观点的研究，同时未对利用语言引导LLM的相关问题展开讨论，此外还忽视了提示词语言对LLM观点对齐的潜在影响。本研究旨在填补这些空白。为此，我们构建了一个基于World Values Survey（世界价值观调查，WVS）的评估框架，系统评估LLM在全球不同国家、不同语言以及不同历史时期下与人类观点的对齐情况。研究发现，LLM仅与少数国家的观点实现了恰当对齐或过度对齐，而与大多数国家的观点存在对齐不足的问题。此外，与现有引导方法相比，将提示词语言调整为与问卷所用语言一致，能够更有效地引导LLM与对应国家的观点对齐。同时，LLM与当代人群的观点对齐程度更高。据我们所知，本研究是首个从全球、语言和时间三个维度对LLM观点对齐主题展开全面探究的研究。我们的代码和数据已公开发布于https://github.com/ku-nlp/global-opinion-alignment和https://github.com/nlply/global-opinion-alignment。
英文摘要：
Today's large language models (LLMs) are capable of supporting multilingual scenarios, allowing users to interact with LLMs in their native languages. When LLMs respond to subjective questions posed by users, they are expected to align with the views of specific demographic groups or historical periods, shaped by the language in which the user interacts with the model. Existing studies mainly focus on researching the opinions represented by LLMs among demographic groups in the United States or a few countries, lacking worldwide country samples and studies on human opinions in different historical periods, as well as lacking discussion on using language to steer LLMs. Moreover, they also overlook the potential influence of prompt language on the alignment of LLMs' opinions. In this study, our goal is to fill these gaps. To this end, we create an evaluation framework based on the World Values Survey (WVS) to systematically assess the alignment of LLMs with human opinions across different countries, languages, and historical periods around the world. We find that LLMs appropriately or over-align the opinions with only a few countries while under-aligning the opinions with most countries. Furthermore, changing the language of the prompt to match the language used in the questionnaire can effectively steer LLMs to align with the opinions of the corresponding country more effectively than existing steering methods. At the same time, LLMs are more aligned with the opinions of the contemporary population. To our knowledge, our study is the first comprehensive investigation of the topic of opinion alignment in LLMs across global, language, and temporal dimensions. Our code and data are publicly available at https://github.com/ku-nlp/global-opinion-alignment and https://github.com/nlply/global-opinion-alignment.
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。
