Empowering Economic Simulation Through Situation-Aware Llm-Driven Generative System

2026-06-16

AI Summary

This study addresses the limitations of traditional macroeconomic models in capturing individual heterogeneity and social interactions, as well as the poor generalization of existing agent-based modeling (ABM) approaches. To overcome these challenges, the authors propose SAMAS, a novel framework that integrates context-aware large language models (LLMs) into economic simulation, endowing agents with macroeconomic reasoning, historical memory, and human-like decision-making capabilities. By synergistically combining ABM, LLM-based role-playing, reinforcement learning reward mechanisms, and macro-trajectory modeling, SAMAS enables the co-evolution of micro-level behaviors and macro-level structures. Experimental results demonstrate that SAMAS significantly outperforms state-of-the-art ABM methods in generating realistic economic fluctuations and predicting turning points, substantially enhancing both out-of-distribution generalization and empirical fidelity.

Abstract

Traditional economic modeling typically follows a TOP-DOWN paradigm, neglecting individual diversity and the complexity of social interactions. To better capture the complexity of societal structure, Agent-Based Modeling (ABM) employs a BOTTOM-UP solution by incorporating micro-level dynamics to generate macroeconomic phenomena. Reinforcement Learning further improves its decision-making ability through tailored reward signals. However, existing ABM systems struggle to generalize beyond predefined scenarios. Recognizing the potential of LLM-driven role-playing in perception and human-like decision-making, we propose SAMAS, which models individual agents with rich macroeconomic understanding embedded in LLMs and economic trajectories experienced in the passing simulation steps. By jointly modeling both macro-level structural patterns and micro-level dynamic behaviors, SAMAS achieves superior performance in volatility realism and turning point prediction.

Problem

• Agent-Based Modeling
• economic simulation
• generalization
• macroeconomic phenomena
• social interactions

Innovation

• LLM-driven agent
• Agent-Based Modeling
• economic simulation
• situation-aware modeling
• macro-micro integration