行为学基础模型 Chiron 发布:预测个体行为误差低于前沿 LLM
A Foundation Model to Measure and Predict Human Behaviour
用 28 万条人物记录微调出的行为模型,预测个体行为比 GPT 这类前沿模型准 28.4%,跑社会科学实验模拟时效应量误差降九成。做用户研究、市场模拟或社科计算研究的团队值得拿它替换通用 LLM 试一下。
研究者发布基础模型 Chiron,专门用于测量和预测人类行为。训练数据覆盖 36 个行为领域、78 项社会科学实验,包含 288,307 条个体记录、1,411,750 个原子行为决策,以及覆盖 109 个心理构念的 99,953 份心理测量评估。
技术上,团队对 LLM 进行联合微调,使其同时完成三件事:表征心理学构念、预测个体行为、将模拟的群体响应分布对齐真实人类统计。评估显示,Chiron 的心理测量结果与人类自评报告更一致,且削弱心理测量效度的偏差更少。
在个体行为预测上,Chiron 比前沿 LLM 最高取得 28.4% 的相对提升;模拟社会科学实验时,效应量估计误差降低 82.2%–94.3%。这意味着用通用 LLM 模拟人群会系统性扭曲统计结果,专门微调的行为模型可显著改善这一点。
A Foundation Model to Measure and Predict Human Behaviour
Large language models (LLMs) are increasingly used to infer psychological attributes, predict individual behaviour, and simulate human populations. However, despite generating plausible responses, they often misrepresent individuals and distort population statistics. Here we introduce Chiron, a foundation model to measure and predict human behaviour. We jointly fine-tune an LLM to represent psychological constructs, predict individual behaviour, and align population response distributions with human statistics. The training and evaluation data span 36 behavioural domains and 78 social science experiments, comprising 288,307 person records, 1,411,750 atomic behavioural decisions, and 99,953 psychometric assessments over 109 psychological constructs. In our evaluations, Chiron's psychological measurements agree more closely with human self-reports and exhibit fewer biases that undermine psychometric validity. It predicts individual behaviour more accurately than frontier LLMs (up to 28.4% relative improvement). In simulations of social science experiments, it more accurately captures the direction and magnitude of human experimental effects (82.2--94.3% lower effect-size estimation error). Together, these results establish Chiron as a unified computational instrument for measuring and predicting human behaviour.
本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:Chiron、基础模型、LLM、心理测量,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Haoran Ye, Tianrui Huo, Hangzhen Li, Xin Zhang, Guojie Song),经 AITOP 多维评分管线结构化提炼。