A Visualized Framework for Event Cooperation with Generative Agents

2025-09-16

AI Summary

Existing LLM-based agent social simulation frameworks lack systematic event evaluation capabilities and fail to integrate with embodied physical environments for visualization, thereby limiting modeling of spatial navigation and object interaction. This paper introduces MiniAgentPro—the first visualization-enabled simulation and evaluation framework designed for event-driven multi-agent collaboration. It integrates an editable 2D map, generative agent control mechanisms, and a dynamic animation player. We construct a standardized benchmark comprising eight progressively complex scenarios to enable intuitive modeling and quantitative assessment of multi-agent spatial coordination. Experimental results show that GPT-4o excels in basic tasks but exhibits significant coordination bottlenecks in higher-order collaborative scenarios. MiniAgentPro advances research on explainable, visualizable, and evaluable embodied multi-agent social behavior.

Technology Category

• Multiagent Systems: Coordination and Collaboration
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• 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
• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Large Language Models (LLMs) have revolutionized the simulation of agent societies, enabling autonomous planning, memory formation, and social interactions. However, existing frameworks often overlook systematic evaluations for event organization and lack visualized integration with physically grounded environments, limiting agents' ability to navigate spaces and interact with items realistically. We develop MiniAgentPro, a visualization platform featuring an intuitive map editor for customizing environments and a simulation player with smooth animations. Based on this tool, we introduce a comprehensive test set comprising eight diverse event scenarios with basic and hard variants to assess agents' ability. Evaluations using GPT-4o demonstrate strong performance in basic settings but highlight coordination challenges in hard variants.

Problem

• Develops a visualization platform for agent event organization
• Addresses lack of systematic evaluation in event cooperation scenarios
• Enables realistic agent navigation and item interaction in environments

Innovation

• Visualization platform with map editor
• Simulation player with smooth animations
• Comprehensive test set for event scenarios