EcoLANG: Efficient and Effective Agent Communication Language Induction for Social Simulation

2025-05-11

AI Summary

Large language models (LLMs) face substantial computational overhead and a trade-off between generalizability and accuracy in social simulation. To address this, we propose EcoLANG—a lightweight agent communication language generation framework tailored for large-scale agent-based modeling (ABM) simulations. Our method introduces natural selection as a novel paradigm for communication language evolution, implemented via a two-stage mechanism: semantic-equivalence word filtering and syntactic rule optimization—enabling simultaneous language compression and evolutionary refinement. Crucially, EcoLANG is the first framework to embed the evolved language directly into the closed-loop simulation. It integrates natural selection algorithms, semantic compression, prompt engineering, and behavioral modeling into an end-to-end evolvable system. Experiments across diverse social simulation scenarios demonstrate that EcoLANG reduces token consumption by over 20%, significantly accelerates inference, and—critically—preserves the original fidelity of social dynamic modeling without any performance compromise.

Technology Category

• Multiagent Systems: Agent Communication
• Natural Language Processing: (Large) Language Models
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Large language models (LLMs) have demonstrated an impressive ability to role-play humans and replicate complex social dynamics. While large-scale social simulations are gaining increasing attention, they still face significant challenges, particularly regarding high time and computation costs. Existing solutions, such as distributed mechanisms or hybrid agent-based model (ABM) integrations, either fail to address inference costs or compromise accuracy and generalizability. To this end, we propose EcoLANG: Efficient and Effective Agent Communication Language Induction for Social Simulation. EcoLANG operates in two stages: (1) language evolution, where we filter synonymous words and optimize sentence-level rules through natural selection, and (2) language utilization, where agents in social simulations communicate using the evolved language. Experimental results demonstrate that EcoLANG reduces token consumption by over 20%, enhancing efficiency without sacrificing simulation accuracy.

Problem

• Reducing high time and computation costs in social simulations
• Addressing inference costs without compromising accuracy and generalizability
• Optimizing agent communication language for efficiency and effectiveness

Innovation

• Evolves language through natural selection optimization
• Reduces token consumption by over 20%
• Enhances efficiency without sacrificing accuracy