When Reasoning Models Hurt Behavioral Simulation: A Solver-Sampler Mismatch in Multi-Agent LLM Negotiation

TL;DR

This paper investigates how large language models' reasoning abilities may hinder their effectiveness in simulating realistic multi-agent negotiations, emphasizing the importance of aligning model evaluation with behavioral objectives.

Contribution

It reveals that stronger reasoning in LLMs can lead to rigid authority decisions in negotiations, highlighting the need to evaluate models based on their behavioral roles rather than strategic prowess.

Findings

Native reasoning models often default to authority-heavy outcomes.

Increased reasoning does not improve behavioral diversity in negotiations.

Structural scaffolds are more effective than reasoning enhancements in promoting negotiated outcomes.

Abstract

Behavioral simulation and strategic problem solving are different tasks. Large language models are increasingly explored as agents in policy-facing institutional simulations, but stronger reasoning need not improve behavioral sampling. We study this solver-sampler mismatch in three multi-agent negotiation environments: two trading-limits scenarios with different authority structures and a grid-curtailment case in emergency electricity management. Across two primary model families, native reasoning and often no reflection collapse toward authority-heavy outcomes. The sharpest case is DeepSeek native reasoning in the grid-curtailment transfer: it reaches action entropy 1.256 and a concession-arc rate of 0.933, yet still ends in authority decision in 15 of 15 runs. A direct OpenAI extension shows the same pressure at provider breadth: GPT-5.2 native reasoning ends in authority decisions in 45 of 45 runs across the three environments. Budget-matched no-reflection controls and orthogonal private-state controls remain rigid, while the negotiation-structured scaffold condition is the only condition that consistently opens negotiated outcomes. These diagnostics are failure screens within a fixed negotiation grammar, not evidence of external behavioral realism or policy-forecasting validity. The results show that neither more output space nor generic extra private state rescues solver-like sampler failure. For institutional simulation, solver strength and sampler qualification are different objectives: models should be evaluated for the behavioral role they are meant to play, not only for strategic capability.