Findings of the Association for Computational Linguistics: EMNLP 2025 2025
Evolution in Simulation: AI-Agent School with Dual Memory for High-Fidelity Educational Dynamics
Corresponding authors: Haoming Wang, Chengliang Wang
DOI 10.18653/v1/2025.findings-emnlp.312
Issue publication: 2025.
Research summary
AI-Agent School models educational interactions through an experience-reflection-optimization cycle and a dual memory architecture. Its experience and knowledge bases combine short- and long-term memory. The work examines how agents evolve through simulated school interactions, providing a computational approach to studying educational dynamics.
Study design.
Computational educational simulation using LLM agents, a Zero-Exp strategy and dual memory.
Abstract
Large language models (LLMs) based Agents are increasingly pivotal in simulating and understanding complex human systems and interactions. We propose the AI-Agent School (AAS) system, built around a self-evolving mechanism that leverages agents for simulating complex educational dynamics. Addressing the fragmented issues in teaching process modeling and the limitations of agents performance in simulating diverse educational participants, AAS constructs the Zero-Exp strategy, employs a continuous “experience-reflection-optimization” cycle, grounded in a dual memory base comprising experience and knowledge bases and incorporating short-term and long-term memory components. Through this mechanism, agents autonomously evolve via situated interactions within diverse simulated school scenarios. This evolution enables agents to more accurately model the nuanced, multi-faceted teacher-student engagements and underlying learning processes found in physical schools. Experiment confirms that AAS can effectively simulate intricate educational dynamics and is effective in fostering advanced agent cognitive abilities, providing a foundational stepping stone from the “Era of Experience” to the “Era of Simulation” by generating high-fidelity behavioral and interaction data.
Research context
How can agent memory and reflection mechanisms support simulation of educational interactions?
Use this work for educational multi-agent simulation, agent memory architecture and synthetic interaction modeling. Its evaluation concerns simulated educational dynamics.
multi-agent simulation · educational dynamics · dual memory · large language models
中文研究介绍
AI-Agent School通过“经验—反思—优化”循环和双记忆架构模拟教育互动，将经验库、知识库与长短期记忆相结合，研究智能体如何在模拟学校情境中演化，为教育动态研究提供计算方法。
Read and cite
Jin, S., Wang, H., Gao, Z., Yang, Y., Chunjia, B., & Wang, C. (2025). Evolution in Simulation: AI-Agent School with Dual Memory for High-Fidelity Educational Dynamics. Findings of the Association for Computational Linguistics: EMNLP 2025, 5843-5857. https://doi.org/10.18653/v1/2025.findings-emnlp.312