生成式智能体模拟城市出行行为与真实调查对齐研究
原标题：Towards Adaptive and Real-Survey-Aligned Simulated Urban Mobility Behaviour of Generative Agents
AI 摘要
该论文构建了法国图卢兹的日常多模式出行智能体模型，对比15种决策者（简单基线、基于调查数据拟合的模型、最小提示与调优提示的生成式智能体、类型化分类器）在真实家庭出行调查上的表现。结果显示，调优提示的生成式智能体在无交通中断的模拟日模态分担上接近拟合模型，但逐次出行层面仍有差距；生成式智能体还能对未预见事件做出反应。新发布的Jev-1.13 TypeSafe分类器表现意外出色，表明生成输出文本并非必要。
正文节选
Towards Adaptive and Real-Survey-Aligned Simulated Urban Mobility Behaviour of Generative Agents Abstract In existing agent-based simulations of multimodal urban mobility, agents, representing humans, choose their travel modes either using discrete choice or machine learning models fitted on structured travel survey data, or with rules explicitly written by the domain experts. Such approaches fix a priori the agents’ possible behaviour space. Hence, a factor absent from variables or rules cannot