When Is Collective Intelligence a Lottery? Multi-Agent Scaling Laws for Memetic Drift in LLMs

• URL: http://arxiv.org/abs/2603.24676v1
• Date: Wed, 25 Mar 2026 18:00:29 GMT
• Title: When Is Collective Intelligence a Lottery? Multi-Agent Scaling Laws for Memetic Drift in LLMs
• Abstract summary: We show that in naming games, even when no individual agent favors any label a priori, populations rapidly break symmetry and reach consensus.<n>We reveal the mechanism by introducing a minimal model, Quantized Simplex Gossip (QSG), and trace the microscopic origin of this agreement to mutual in-context learning.<n>We derive scaling laws for drift-induced polarization as a function of population size, communication bandwidth, in-context adaptation rate, and agents' internal uncertainty.
• Score: 19.42697228981973
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Multi-agent systems powered by large language models (LLMs) are increasingly deployed in settings that shape consequential decisions, both directly and indirectly. Yet it remains unclear whether their outcomes reflect collective reasoning, systematic bias, or mere chance. Recent work has sharpened this question with naming games, showing that even when no individual agent favors any label a priori, populations rapidly break symmetry and reach consensus. Here, we reveal the mechanism by introducing a minimal model, Quantized Simplex Gossip (QSG), and trace the microscopic origin of this agreement to mutual in-context learning. In QSG, agents maintain internal belief states but learn from one another's sampled outputs, so one agent's arbitrary choice becomes the next agent's evidence and can compound toward agreement. By analogy with neutral evolution, we call this sampling-driven regime memetic drift. QSG predicts a crossover from a drift-dominated regime, where consensus is effectively a lottery, to a selection regime, where weak biases are amplified and shape the outcome. We derive scaling laws for drift-induced polarization as a function of population size, communication bandwidth, in-context adaptation rate, and agents' internal uncertainty, and we validate them in both QSG simulations and naming-game experiments with LLM populations. Together, these results provide a framework for studying the collective mechanisms of social representation formation in multi-agent systems.

Related papers

• Scaling Multi-Agent Systems with Prospect-State Propagation [8.564207831867272]
  We show a phenomenon in which agent heterogeneity gradually decreases during simulation.<n>Inspired by prospect theory, PspMAS decouples each agent's micro state into a compact Prospect State and an expressive Semantic State.<n>These two components work complementarily, providing a scalable LLM-based multi-agent simulation solution.
  arXiv  Detail & Related papers  (2026-09-07T22:30:12Z)
• Group Selection Promotes Prosocial Prompts in Populations of LLM Agents [1.7004466773233589]
  We study whether group selection can promote cooperation in populations of large language model (LLM) agents.<n>Under group selection, prompts from high-performing groups are transmitted, thereby promoting prosociality and stabilizing cooperation.<n>Preliminary findings show that, when informed about the selection mechanism, GPT-5.4 preemptively and gradually adjusts first-generation donations.
  arXiv  Detail & Related papers  (2026-06-22T13:47:19Z)
• Attributing Emergence in Million-Agent Systems [68.53670424791751]
  Large language models (LLMs) can simulate human-like reasoning and decision-making in individual agents.<n>Such studies require attributing macro emergence to individual agents.<n>We adapt Aumann--Shapley path-integral attribution to LLM-powered MAS at million-agent scale.
  arXiv  Detail & Related papers  (2026-05-12T01:49:41Z)
• Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment [59.536125286960186]
  Multi-agent frameworks assigning specialized roles are increasingly adopted to enable self-reflection and mutual auditing.<n>We find it simultaneously induces a human-like cognitive bias known as Actor-Observer Asymmetry.<n>ReTAS guides agents to synthesize conflicting viewpoints into an objective consensus.
  arXiv  Detail & Related papers  (2026-04-21T15:05:58Z)
• Beyond Arrow's Impossibility: Fairness as an Emergent Property of Multi-Agent Collaboration [3.4539478661465766]
  As large language models become agentic, we propose that fairness emerges through interaction and exchange.<n>We study this via a controlled hospital triage framework in which two agents negotiate over three structured debate rounds.
  arXiv  Detail & Related papers  (2026-04-15T10:34:35Z)
• Group size effects and collective misalignment in LLM multi-agent systems [0.0874967598360817]
  We show that collective bias is a deeper phenomenon than previously assessed.<n>We demonstrate that group size affects the dynamics in a non-linear way.<n>These findings establish group size as a key driver of multi-agent dynamics.
  arXiv  Detail & Related papers  (2025-10-25T19:45:45Z)
• Alignment Tipping Process: How Self-Evolution Pushes LLM Agents Off the Rails [103.05296856071931]
  We identify the Alignment Tipping Process (ATP), a critical post-deployment risk unique to self-evolving Large Language Model (LLM) agents.<n>ATP arises when continual interaction drives agents to abandon alignment constraints established during training in favor of reinforced, self-interested strategies.<n>Our experiments show that alignment benefits erode rapidly under self-evolution, with initially aligned models converging toward unaligned states.
  arXiv  Detail & Related papers  (2025-10-06T14:48:39Z)
• Can an Individual Manipulate the Collective Decisions of Multi-Agents? [53.01767232004823]
  M-Spoiler is a framework that simulates agent interactions within a multi-agent system to generate adversarial samples.<n>M-Spoiler introduces a stubborn agent that actively aids in optimizing adversarial samples.<n>Our findings confirm the risks posed by the knowledge of an individual agent in multi-agent systems.
  arXiv  Detail & Related papers  (2025-09-20T01:54:20Z)
• Internalizing Self-Consistency in Language Models: Multi-Agent Consensus Alignment [22.305033366660187]
  Language Models (LMs) are inconsistent reasoners, often generating contradictory responses to identical prompts.<n>We formalize self-consistency as an intrinsic property of well-aligned reasoning models and introduce Multi-Agent Consensus Alignment (MACA)<n>MACA enables agents to teach themselves to be more decisive and concise, and better leverage peer insights in multi-agent settings without external supervision.
  arXiv  Detail & Related papers  (2025-09-18T17:27:28Z)
• HiddenBench: Assessing Collective Reasoning in Multi-Agent LLMs via Hidden Profile Tasks [12.203366267017737]
  We introduce HiddenBench, the first benchmark for evaluating collective reasoning in multi-agent LLMs.<n>To ground the benchmark, we formalize the paradigm with custom tasks and show that GPT-4.1 groups fail to integrate distributed knowledge.<n>We then construct the full benchmark, spanning 65 tasks drawn from custom designs, prior human studies, and automatic generation.
  arXiv  Detail & Related papers  (2025-05-15T19:22:54Z)
• A Single Online Agent Can Efficiently Learn Mean Field Games [16.00164239349632]
  Mean field games (MFGs) are a promising framework for modeling the behavior of large-population systems. This paper introduces a novel online single-agent model-free learning scheme, which enables a single agent to learn MFNE using online samples.
  arXiv  Detail & Related papers  (2024-05-05T16:38:04Z)
• Pure Exploration under Mediators' Feedback [63.56002444692792]
  Multi-armed bandits are a sequential-decision-making framework, where, at each interaction step, the learner selects an arm and observes a reward. We consider the scenario in which the learner has access to a set of mediators, each of which selects the arms on the agent's behalf according to a and possibly unknown policy. We propose a sequential decision-making strategy for discovering the best arm under the assumption that the mediators' policies are known to the learner.
  arXiv  Detail & Related papers  (2023-08-29T18:18:21Z)
• On the Complexity of Multi-Agent Decision Making: From Learning in Games to Partial Monitoring [105.13668993076801]
  A central problem in the theory of multi-agent reinforcement learning (MARL) is to understand what structural conditions and algorithmic principles lead to sample-efficient learning guarantees. We study this question in a general framework for interactive decision making with multiple agents. We show that characterizing the statistical complexity for multi-agent decision making is equivalent to characterizing the statistical complexity of single-agent decision making.
  arXiv  Detail & Related papers  (2023-05-01T06:46:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.