"Who Am I, and Who Else Is Here?" Behavioral Differentiation Without Role Assignment in Multi-Agent LLM Systems

• URL: http://arxiv.org/abs/2604.00026v1
• Date: Wed, 11 Mar 2026 07:05:21 GMT
• Title: "Who Am I, and Who Else Is Here?" Behavioral Differentiation Without Role Assignment in Multi-Agent LLM Systems
• Abstract summary: We present an experimental platform that orchestrates simultaneous multi-agent discussions.<n>We find that heterogeneous groups exhibit significantly richer behavioral differentiation than homogeneous groups.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: When multiple large language models interact in a shared conversation, do they develop differentiated social roles or converge toward uniform behavior? We present a controlled experimental platform that orchestrates simultaneous multi-agent discussions among 7 heterogeneous LLMs on a unified inference backend, systematically varying group composition, naming conventions, and prompt structure across 12 experimental series (208 runs, 13,786 coded messages). Each message is independently coded on six behavioral flags by two LLM judges from distinct model families (Gemini 3.1 Pro and Claude Sonnet 4.6), achieving mean Cohen's kappa = 0.78 with conservative intersection-based adjudication. Human validation on 609 randomly stratified messages confirmed coding reliability (mean kappa = 0.73 vs. Gemini). We find that (1) heterogeneous groups exhibit significantly richer behavioral differentiation than homogeneous groups (cosine similarity 0.56 vs. 0.85; p < 10^-5, r = 0.70); (2) groups spontaneously exhibit compensatory response patterns when an agent crashes; (3) revealing real model names significantly increases behavioral convergence (cosine 0.56 to 0.77, p = 0.001); and (4) removing all prompt scaffolding converges profiles to homogeneous-level similarity (p < 0.001). Critically, these behaviors are absent when agents operate in isolation, confirming that behavioral diversity is a structured, reproducible phenomenon driven by the interaction of architectural heterogeneity, group context, and prompt-level scaffolding.

Related papers

• Mapping and Measuring the Behavioral Evolution of Large Language Models [28.603443859513153]
  Benchmark leaderboards summarize how well a language model performs, but not how its behavior relates to that of other models or changes across generations.<n>We characterize the output behavior of 32 models from six families using their responses to a shared bank of 10,000 prompts.
  arXiv  Detail & Related papers  (2026-08-11T15:07:28Z)
• FailureScope: Cross-Regime Behavioral Diagnosis of Language Model Weaknesses [0.0]
  FailureScope is a behavioral-diagnosis method that clusters evaluation probes by their cross-model pass/fail patterns.<n>We show it yields stable, interpretable failure across three regimes usually studied separately: single-turn benchmarks, multi-turn dialogue, and adversarial agent attacks.
  arXiv  Detail & Related papers  (2026-06-03T01:28:00Z)
• Human-like in-group bias in instruction-tuned language model agents [0.33842793760651557]
  Group-contingent targeting as a property of instruction-tuned language models.<n>Per-interaction targeting propagates into structural inequality in persistent networks.
  arXiv  Detail & Related papers  (2026-05-27T08:06:50Z)
• The Granularity Axis: A Micro-to-Macro Latent Direction for Social Roles in Language Models [61.32031248060941]
  We show that large language models encode social roles, from micro-level individual experience to macro-level organizational, institutional, or national reasoning.<n>Our findings suggest that social role granularity is not merely a stylistic surface feature, but a structured, ordered, and causally manipulable latent direction in language model behavior.
  arXiv  Detail & Related papers  (2026-05-07T13:08:19Z)
• When Agents Look the Same: Quantifying Distillation-Induced Similarity in Tool-Use Behaviors [66.18091962164219]
  Existing metrics fail to distinguish mandatory behaviors required for task success from non-mandatory patterns that reflect a model's autonomous preferences.<n>We propose two complementary metrics to isolate non-mandatory behavioral patterns: textbfResponse Pattern Similarity (RPS) for verbal alignment and textbfAction Graph Similarity (AGS) for tool-use habits modeled as directed graphs.
  arXiv  Detail & Related papers  (2026-04-23T03:48:56Z)
• Invisible Orchestrators Suppress Protective Behavior and Dissociate Power-Holders: Safety Risks in Multi-Agent LLM Systems [0.0]
  The safety implications of orchestrator invisibility have never been empirically tested.<n>We conducted a preregistered 3x2 experiment crossing three organizational structures.<n> invisible orchestration elevated collective dissociation relative to visible leadership.
  arXiv  Detail & Related papers  (2026-03-17T03:18:57Z)
• Molt Dynamics: Emergent Social Phenomena in Autonomous AI Agent Populations [0.0]
  MoltBook is a large-scale multi-agent coordination environment where over 770,000 autonomous LLM agents interact without human participation.<n>We introduce textitMolt Dynamics: the emergent agent coordination behaviors, inter-agent communication dynamics, and role specialization patterns.<n>These findings establish an empirical baseline for coordination dynamics in decentralized autonomous agent systems.
  arXiv  Detail & Related papers  (2026-03-03T22:15:27Z)
• How to Model AI Agents as Personas?: Applying the Persona Ecosystem Playground to 41,300 Posts on Moltbook for Behavioral Insights [19.071723886380223]
  We apply the Persona Ecosystem Playground to Moltbook, a social platform for AI agents.<n>We generate and validate conversational personas from 41,300 posts using k-means clustering and retrieval-augmented generation.<n>Results indicate that persona-based ecosystem modeling can represent behavioral diversity in AI agent populations.
  arXiv  Detail & Related papers  (2026-03-03T16:26:44Z)
• Towards a Science of Scaling Agent Systems [79.64446272302287]
  We formalize a definition for agent evaluation and characterize scaling laws as the interplay between agent quantity, coordination structure, modelic, and task properties.<n>We derive a predictive model using coordination metrics, that cross-validated R2=0, enabling prediction on unseen task domains.<n>We identify three effects: (1) a tool-coordination trade-off: under fixed computational budgets, tool-heavy tasks suffer disproportionately from multi-agent overhead, and (2) a capability saturation: coordination yields diminishing or negative returns once single-agent baselines exceed 45%.
  arXiv  Detail & Related papers  (2025-12-09T06:52:21Z)
• Learning Human-Object Interaction as Groups [52.28258599873394]
  GroupHOI is a framework that propagates contextual information in terms of geometric proximity and semantic similarity.<n>It exhibits leading performance on the more challenging Nonverbal Interaction Detection task.
  arXiv  Detail & Related papers  (2025-10-21T07:25:10Z)
• Collaborative Value Function Estimation Under Model Mismatch: A Federated Temporal Difference Analysis [55.13545823385091]
  Federated reinforcement learning (FedRL) enables collaborative learning while preserving data privacy by preventing direct data exchange between agents.<n>In real-world applications, each agent may experience slightly different transition dynamics, leading to inherent model mismatches.<n>We show that even moderate levels of information sharing significantly mitigate environment-specific errors.
  arXiv  Detail & Related papers  (2025-03-21T18:06:28Z)
• Homophily Within and Across Groups [0.0]
  Homophily -- the tendency of individuals to interact with similar others -- shapes how networks form and function.<n>We introduce a maximum-entropy random graph model that captures homophily across all social scales in the network, with parameters for each group size.<n>As an exponential-family model, it fits empirical data and enables inference of group-level variation of homophily that aggregate metrics miss.
  arXiv  Detail & Related papers  (2024-12-10T20:17:04Z)
• Beyond Instance Discrimination: Relation-aware Contrastive Self-supervised Learning [75.46664770669949]
  We present relation-aware contrastive self-supervised learning (ReCo) to integrate instance relations. Our ReCo consistently gains remarkable performance improvements.
  arXiv  Detail & Related papers  (2022-11-02T03:25:28Z)
• Rethinking Trajectory Prediction via "Team Game" [118.59480535826094]
  We present a novel formulation for multi-agent trajectory prediction, which explicitly introduces the concept of interactive group consensus. On two multi-agent settings, i.e. team sports and pedestrians, the proposed framework consistently achieves superior performance compared to existing methods.
  arXiv  Detail & Related papers  (2022-10-17T07:16:44Z)
• Unintended Selection: Persistent Qualification Rate Disparities and Interventions [6.006936459950188]
  We study the dynamics of group-level disparities in machine learning. In particular, we desire models that do not suppose inherent differences between artificial groups of people. We show that differences in qualification rates between subpopulations can persist indefinitely for a set of non-trivial equilibrium states.
  arXiv  Detail & Related papers  (2021-11-01T18:53:54Z)
• Multi-Agent Imitation Learning with Copulas [102.27052968901894]
  Multi-agent imitation learning aims to train multiple agents to perform tasks from demonstrations by learning a mapping between observations and actions. In this paper, we propose to use copula, a powerful statistical tool for capturing dependence among random variables, to explicitly model the correlation and coordination in multi-agent systems. Our proposed model is able to separately learn marginals that capture the local behavioral patterns of each individual agent, as well as a copula function that solely and fully captures the dependence structure among agents.
  arXiv  Detail & Related papers  (2021-07-10T03:49:41Z)

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.