Negotiating Comfort: Simulating Personality-Driven LLM Agents in Shared Residential Social Networks

2025-07-13

AI Summary

This study investigates how residents’ personality-driven behaviors influence collective decision-making on central heating temperature in shared housing. We propose a generative multi-agent simulation framework powered by large language models (LLMs), implementing a two-tier negotiation mechanism between household members and resident representatives. The model integrates the Big Five personality trait distributions, empirically grounded social network structures, and environmental constraints to simulate daily temperature consensus formation. Our key contribution lies in the first systematic coupling of LLM agents’ personality-augmented reasoning capabilities with social-psychological variables—decisiveness, altruism, and sociability—to quantify their effects on individual well-being and group-level outcomes. Experimental results identify temperature preference, decisiveness, and altruism as primary predictors; positively oriented traits significantly enhance both subjective well-being and negotiation efficiency—demonstrating LLMs’ validity and explanatory power in simulating complex socio-behavioral dynamics.

Technology Category

• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Humans and AI: Learning Human Values and Preferences

Application Category

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

Abstract

We use generative agents powered by large language models (LLMs) to simulate a social network in a shared residential building, driving the temperature decisions for a central heating system. Agents, divided into Family Members and Representatives, consider personal preferences, personal traits, connections, and weather conditions. Daily simulations involve family-level consensus followed by building-wide decisions among representatives. We tested three personality traits distributions (positive, mixed, and negative) and found that positive traits correlate with higher happiness and stronger friendships. Temperature preferences, assertiveness, and selflessness have a significant impact on happiness and decisions. This work demonstrates how LLM-driven agents can help simulate nuanced human behavior where complex real-life human simulations are difficult to set.

Problem

• Simulating personality-driven LLM agents in shared residential networks
• Modeling temperature decisions based on social and weather factors
• Analyzing impact of personality traits on happiness and decisions

Innovation

• LLM-driven agents simulate residential social networks
• Agents consider traits, preferences, and weather
• Positive traits correlate with higher happiness