arXiv cs.CL· Zhengyu Hu, Jianxun Lian, Zheyuan Xiao, Max Xiong, Tianfu Wang, Teng Xiao, Yuxuan Lei, Fengqing Jiang, Kaize Ding, Ziang Xiao, Nicholas Jing Yuan, Xing Xie, Radha Poovendran·· 11 小时前AI 评分31
面向 LLM 社会模拟的群体对齐人设生成框架
Population-Aligned Persona Generation for LLM-based Social Simulation
AI 导读
研究者提出一套面向 LLM 社会模拟的群体对齐人设生成框架，先用 LLM 从长期社交媒体数据生成叙事型人设并做质量评估过滤低保真档案，再通过重要性采样对齐 Big Five 等参考心理测量分布。该框架还引入任务特定模块，将全局对齐的人设集适配到目标子群体。实验显示该方法显著降低群体层面偏差，支持更准确灵活的社会模拟。
正文 · 原文
Authors:Zhengyu Hu, Jianxun Lian, Zheyuan Xiao, Max Xiong, Tianfu Wang, Teng Xiao, Yuxuan Lei, Fengqing Jiang, Kaize Ding, Ziang Xiao, Nicholas Jing Yuan, Xing Xie, Radha Poovendran
Abstract:Recent advances in large language models (LLMs) have enabled large-scale, high-fidelity social simulations, creating new opportunities for computational social science. However, constructing persona sets that faithfully reflect real-world population diversity remains a key challenge. Existing studies often emphasize agentic frameworks and simulation environments, while paying less attention to persona generation and the biases introduced by unrepresentative persona sets. In this paper, we propose a systematic framework for synthesizing high-quality, population-aligned persona sets for LLM-driven social simulation. Our approach begins by leveraging LLMs to generate narrative personas from long-term social media data, followed by rigorous quality assessment to filter out low-fidelity profiles. We then apply importance sampling to achieve global alignment with reference psychometric distributions, such as the Big Five personality traits. To address the needs of specific simulation contexts, we further introduce a task-specific module that adapts the globally aligned persona set to targeted subpopulations. Extensive experiments demonstrate that our method significantly reduces population-level bias and enables accurate, flexible social simulation for a wide range of research and policy applications.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2509.10127 [cs.CL] |
| (or arXiv:2509.10127v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2509.10127 arXiv-issued DOI via DataCite |
Submission history
From: Zhengyu Hu [view email]
[v1]
Fri, 12 Sep 2025 10:43:47 UTC (1,711 KB)
[v2]
Sat, 4 Oct 2025 09:04:38 UTC (1,711 KB)
[v3]
Wed, 4 Nov 2026 09:22:24 UTC (2,031 KB)
来源：arXiv cs.CL · arxiv.org