持续异质性：大语言模型驱动交通中的一种涌现集体机制
Sustained Heterogeneity: an emergent collective mechanism in LLM-driven traffic
- College of Civil Engineering, Nanjing Tech University（南京工业大学土木工程学院）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本研究在LLM控制的交通中识别出持续异质性这一新兴集体机制，通过实验得出密度相关的关键LLM渗透率p_c，表明需在动力学层强制执行稳定性。
AI中文摘要：
大语言模型（LLMs）正越来越多地被用作物理多智能体系统中的闭环控制器，但其涌现的集体动力学仍未得到充分表征。我们按照Sugiyama 2008年的范式，在230米环形道路上部署了22个LLM智能体作为直接的实时目标速度控制器（每0.5秒一个周期，以IDM作为碰撞避免约束），重现了类人的走走停停波。我们系统排除了六种匹配的对照情况，涵盖随机性（白噪声、OU噪声、温度）、群体方差和动力学不稳定性（延迟、OV模型）。幸存的现象被称为持续异质性（Sustained Heterogeneity, SH），即LLM选择的目标速度调整中持续存在的、近似与温度无关的现象（在6倍温度扫描范围内约为8%），该调整通过漂移、间隙侵蚀和非线性制动的三级级联传播。在四种交通密度下，关键LLM渗透率p_c从密度43.5辆/公里时无转变单调下降至密度95.7辆/公里时的p_c约0.23，这与由触发距离、随机性和车队规模控制的启动阈值模型一致。对三个随机种子下的39600个决策进行的思维链分析显示，智能体进行多因素安全推理，但系统性分歧持续存在，这意味着必须在动力学层强制执行稳定性。本研究首次在LLM控制的交通中识别出一种此前未表征的集体机制，并绘制了密度相关的相边界p_c(ρ)。
英文摘要：
Large language models (LLMs) are increasingly adopted as closed-loop controllers in physical multi-agent systems, yet their emergent collective dynamics remain incompletely characterised. We deploy 22 LLM agents as direct, real-time target-speed controllers (per 0.5 s cycle, with IDM as collision-avoidance clamp) on a 230 m ring road under the Sugiyama 2008 paradigm, reproducing human-like stop-and-go waves. Six matched controls spanning stochasticity (white noise, OU noise, temperature), population variance, and dynamical instability (delay, OV model) are systematically excluded. The surviving phenomenon, termed Sustained Heterogeneity (SH), is the persistent, approximately temperature-insensitive (approx. 8 percent across a 6x T sweep), per-cycle divergence in LLM-chosen target-speed adjustments, propagating through a three-stage cascade of drift, gap erosion, and nonlinear braking. Across four traffic densities, the critical LLM penetration fraction p_c decreases monotonically from no transition at density 43.5 veh/km to p_c approx 0.23 at density 95.7 veh/km, consistent with an initiation-threshold model governed by trigger distance, stochasticity, and fleet size. Chain-of-thought analysis of 39,600 decisions across three seeds shows agents engage in multi-factor safety reasoning, yet systematic divergence persists, implying stability must be enforced at the dynamics layer. This is the first study to identify a previously uncharacterised collective mechanism in LLM-controlled traffic and map a density-dependent phase boundary p_c(rho).
