Think-Before-Speak: From Internal Evaluation to Public Expression in Multi-Agent Social Simulation

2026-06-02

AI Summary

Current large language model–driven multi-agent dialogue simulations struggle to capture the full psychological process through which agents transition from internal evaluation to public expression. This work proposes the Thought-Behavior-Speech (TBS) framework, which explicitly models the separation between private reasoning and public utterances for the first time. The framework dynamically updates structured internal states—such as cognitive dissonance and perceived opinion climate—over time steps, while a coordinator modulates speaking intentions and silence behaviors. This design enables the co-evolution of internal assessment and public interaction, rendering micro-level psychological processes observable and analyzable. In simulations of climate policy discussions, TBS generates coherent internal state trajectories, demonstrating that cognitive dissonance increases willingness to speak, perceived pressure to remain silent suppresses expression, and actual public utterances are primarily governed by speaking allocation rules.

Abstract

LLM-based multi-agent simulation offers a promising way to study social interaction, deliberation, and collective opinion dynamics. However, many existing dialogue simulation frameworks represent interaction mainly as observable turn exchange or aggregated outputs, leaving the internal evaluative processes behind silence, speaking intention, and public expression difficult to examine. We introduce TBS (Think-Before-Speak), an interval-based multi-agent simulation framework that separates agents' private reasoning from public utterance generation. At each interval, all agents update structured internal states based on the shared dialogue history and their own memory. These states include dissonance-related appraisal, perceived opinion climate, perceived isolation risk, response strategy, and willingness to speak. The orchestrator then resolves competing speaking intentions and commits one utterance to the public dialogue, allowing internal evaluation and public interaction to co-evolve over time. We evaluate TBS in simulated town hall discussions on a climate-related policy issue. Results show that TBS produces coherent internal-state traces and that these traces vary systematically across turn-allocation, silence, and memory conditions. Dissonance-related appraisal increases agents' willingness to speak, whereas silence-pressure appraisal decreases it. Once speaking intention is formed, public expression is shaped mainly by turn-allocation rules. These findings suggest that TBS supports mechanism-sensitive social simulation by making the pathway from internal evaluation to public expression observable and analyzable.

Problem

• multi-agent simulation
• internal evaluation
• public expression
• dialogue modeling
• social interaction

Innovation

• multi-agent simulation
• internal evaluation
• public expression
• structured internal states
• LLM-based social dynamics