研究大模型如何根据个人与集体信息更新预期，发现其存在明显行为偏差。
Bridging Expectation Signals: LLM-Based Experiments and a Behavioral Kalman Filter Framework
- 构建行为卡尔曼滤波框架，量化大模型对不同信息的权重分配。
- 大模型更依赖个人信号，多重信息源会降低单个信号的权重。
- 企业高管角色的大模型比家庭角色更具系统性偏差，适合做决策研究。
随着大语言模型（LLM）越来越多地扮演经济主体角色，其如何利用异质性信息更新信念的机制仍不明确。本文设计实验并提出行为卡尔曼滤波框架，量化基于大模型的代理（作为家庭或企业首席执行官）在接收到个体与总体信息时的预期更新过程。实验与模型估计结果揭示四个一致模式：(1) 代理对先验与信号的加权偏离单位值；(2) 家庭与企业高管代理均对个体信号赋予远高于总体信号的权重；(3) 多个信息源并存时存在显著负向交互效应，即多重信息削弱了每个单独信号的边际权重；(4) 家庭与企业高管角色的预期形成模式存在显著差异。最后，我们发现LoRA微调可缓解但无法完全消除大模型在预期形成中的行为偏差。
原文摘要 · Abstract (English)
As LLMs increasingly function as economic agents, the specific mechanisms LLMs use to update their belief with heterogeneous signals remain opaque. We design experiments and develop a Behavioral Kalman Filter framework to quantify how LLM-based agents update expectations, acting as households or firm CEOs, update expectations when presented with individual and aggregate signals. The results from experiments and model estimation reveal four consistent patterns: (1) agents' weighting of priors and signals deviates from unity; (2) both household and firm CEO agents place substantially larger weights on individual signals compared to aggregate signals; (3) we identify a significant and negative interaction between concurrent signals, implying that the presence of multiple information sources diminishes the marginal weight assigned to each individual signal; and (4) expectation formation patterns differ significantly between household and firm CEO agents. Finally, we demonstrate that LoRA fine-tuning mitigates, but does not fully eliminate, behavioral biases in LLM expectation formation.
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。