Microscopic dynamics of consensus formation in multi-agent LLM Naming Games
On the map: LLM agents and tool use → Social simulation and multi-agent debate
Abstract (arXiv)
Decentralized populations of Large Language Model (LLM) agents can spontaneously reach consensus on shared conventions, yet the microscopic mechanisms by which their internal stochasticity shapes macroscopic ordering remain unexplored. We study a minimal LLM Naming Game in which the listener's decision is a single-token LLM call at decoding temperature T, replacing the inventory check of the deterministic Naming Game. Each interaction decomposes into an in-inventory and an out-inventory channel with conditional rates π(T)!equiv!P(YESmid w∈ Pj) and φ(T)!equiv!P(YESmid wnotin Pj), whose balance controls an ordering-disordering drift. A mean-field theory of the two-rate dynamics yields an analytical ordering condition that generalizes the consensus threshold of the stochastic Naming Game to a critical line in the (π,φ) plane. Across three open-weight architectures, consensus is always reached, but through three distinct listener regimes: permissive (repaint-noise dominated), near-deterministic, and conservative (missed-collapse dominated). The effective finite-size exponent β(T) in trm conv!~!Nβ shifts with temperature, and the temperature-sensitivity α in tc!~!eα T ranges from ≈,0.67 to ≈,0 across architectures. Decoding temperature thus emerges as an architecture-dependent control parameter for decentralized LLM populations, quantitatively characterized by the statistical-physics toolkit.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use. Builds on 14 of its 17 references that are in the map. First posted on arXiv 3 Aug 2026.
Builds on
14 of 17 references.
Closest ancestors · 7
Same direction, newest first
- 2026Conformity Generates Collective Misalignment in AI Agents Societies11 May
- 2026Collective Alignment in LLM Multi-Agent Systems: Disentangling Bias from Cooperation via Statistical Physics11 May
- 2026Increasing intelligence in AI agents can worsen collective outcomes12 Mar
- 2026Conformity and Social Impact on AI Agents8 Jan
- 2025Group size effects and collective misalignment in LLM multi-agent systems25 Oct
- 2024Cooperate or Collapse: Emergence of Sustainable Cooperation in a Society of LLM Agents25 Apr
- 2023Simulating Opinion Dynamics with Networks of LLM-based Agents16 Nov
Other references · 7
- 2019The Curious Case of Neural Text Degenerationcited by 2,881
- 2024When LLMs Play the Telephone Game: Cultural Attractors as Conceptual Tools to Evaluate LLMs in Multi-turn Settingscited by 17
- 2025Detailed balance in large language model-driven agentscited by 6
- 2025Learning Individual Behavior in Agent-Based Models with Graph Diffusion Networkscited by 4
- 2025Reaching Agreement Among Reasoning LLM Agentscited by 3
- 2026Exploring the Topology and Memory of Consensus: How LLM Agents Agree, Fragment, or Settle When Forming Conventionscited by 2
- 2025SIGN: Schema-Induced Games for Namingcited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.
Similar, not cited
Close in wording, no citation either way
- 2026When Is Collective Intelligence a Lottery? Multi-Agent Scaling Laws for Memetic Drift in LLMs
- 2025Emergent Coordination in Multi-Agent Language Models
- 2026DICE: Entropy-Regularized Equilibrium Selection for Stable Multi-Agent LLM Coordination
- 2026Emergence of Biased Consensus in Multi-Agent LLM Debates
- 2026Coalition Formation in LLM Agent Networks: Stability Analysis and Convergence Guarantees
- 2026Conformity Dynamics in LLM Multi-Agent Systems: The Roles of Topology and Self-Social Weighting
- 2026From Signals to Structure: How Memory Architecture Drives Language Emergence in LLM Agents
- 2026Collective cooperation without individual fidelity in LLM agents