Dynamics of Cognitive Heterogeneity: Investigating Behavioral Biases in Multi-Stage Supply Chains with LLM-Based Simulation

2026-04-18

AI Summary

Traditional supply chain simulations struggle to effectively capture coordination failures arising from human cognitive biases, owing to limitations in scalability and controllability. This study proposes a novel multi-agent simulation framework that systematically integrates large language models—such as GPT and DeepSeek—with heterogeneous reasoning capabilities. By incorporating hierarchical reasoning mechanisms, statistical validation, and behavioral experiments, the framework models dynamic decision-making behaviors across multi-stage interactions. The findings reveal that cognitive heterogeneity commonly induces myopic and self-interested behaviors, thereby exacerbating systemic inefficiencies; however, information sharing significantly mitigates these adverse effects. This work offers a new pathway for understanding how cognitive biases influence collective efficiency and provides theoretical foundations for AI-augmented organizational coordination design.

Technology Category

• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Multiagent Systems: Coordination and Collaboration
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Modeling coordination among generative agents in complex multi-round decision-making presents a core challenge for AI and operations management. Although behavioral experiments have revealed cognitive biases behind supply chain inefficiencies, traditional methods face scalability and control limitations. We introduce a scalable experimental paradigm using Large Language Models (LLMs) to simulate multi-stage supply chain dynamics. Grounded in a Hierarchical Reasoning Framework, this study specifically analyzes the impact of cognitive heterogeneity on agent interactions. Unlike prior homogeneous settings, we employ DeepSeek and GPT agents to systematically vary reasoning sophistication across supply chain tiers. Through rigorously replicated and statistically validated simulations, we investigate how this cognitive diversity influences collective outcomes. Results indicate that agents exhibit myopic and self-interested behaviors that exacerbate systemic inefficiencies. However, we demonstrate that information sharing effectively mitigates these adverse effects. Our findings extend traditional behavioral methods and offer new insights into the dynamics of AI-enabled organizations. This work underscores both the potential and limitations of LLM-based agents as proxies for human decision-making in complex operational environments.

Problem

• cognitive heterogeneity
• behavioral biases
• multi-stage supply chains
• agent interactions
• systemic inefficiencies

Innovation

• LLM-based simulation
• cognitive heterogeneity
• multi-stage supply chain
• hierarchical reasoning framework
• generative agents