发表机构
Tianli Qiming AI Research Institute; Sichuan Qiming Daren Technology Co., Ltd.; Swiss AI Laboratories（天立启明人工智能研究院; 四川启铭达人科技有限公司; 瑞士人工智能实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究基于297915名学习者的自适应辅导历史，构建人机学习系统的动力学模型，揭示可复现的宏观动力学规律，为理解人机协同学习的行为机制提供了可解释的有效场框架。
AI 中文摘要
闭环人机系统会产生高维行为轨迹，其集体动力学仍不明确。研究利用297915名学习者的自适应辅导历史，在模型拟合前定义语义序变量，并在用户不重叠的队列中对其进行测试。该状态呈现可复现的盆状流和经操作定义的、状态异质的类亚稳态动力学。构造匹配的零模型可区分归一化记忆松弛与可复现的过剩场。一个四项条件机制可恢复群体漂移（相关系数r=0.946，学习者自举95%置信区间为0.935-0.955）。预测性事件级自监督学习可恢复状态和学习平面流；零模型参考修正保留了定向、部分振幅的过剩场结构，无需完全校准。打乱顺序的训练会反转有序轨迹上的学习平面流；支持对齐随机化选择性降低向内传输。在无状态监督下，两个轴仍保持线性可及。在跨模型拟合时，模型共享主导群体漂移（r=0.866，学习者自举95%置信区间为0.857-0.875）和持久性排序；残差方向仍为模型特定的。这些结果确定了一个外部锚定的主导阶有效场，其关联了经验动力学、可解释机制和神经计算。
英文摘要
Closed-loop human-AI systems generate high-dimensional behavioural trajectories whose collective dynamics remain obscure. Using 297,915 learners' adaptive-tutoring histories, we define semantic order variables before model fitting and test them in user-disjoint cohorts. The state exhibits reproducible basin-like flow and operationally defined, state-heterogeneous metastable-like kinetics. A construction-matched null distinguishes normalised-memory relaxation from a reproducible excess field. A four-term conditional mechanism recovers population drift (r = 0.946; learner-bootstrap 95% CI, 0.935-0.955). Predictive event-level self-supervised learning recovers the state and learned-plane flow; null-referenced corrections retain directional, partial-amplitude excess-field structure without full calibration. Shuffled-order training reverses learned-plane flow on ordered trajectories; support-alignment randomisation selectively reduces inward transport. Both axes remain linearly accessible without state supervision. Without cross-model fitting, the models share leading population drift (r = 0.866; learner-bootstrap 95% CI, 0.857-0.875) and persistence ordering; residual directions remain model-specific. These results identify an externally anchored leading-order effective field linking empirical dynamics, an interpretable mechanism and neural computation.
Comments8 figures, 11 supplementary tables
