The Chameleon's Limit: Investigating Persona Collapse and Homogenization in Large Language Models

• URL: http://arxiv.org/abs/2604.24698v1
• Date: Mon, 27 Apr 2026 17:01:48 GMT
• Title: The Chameleon's Limit: Investigating Persona Collapse and Homogenization in Large Language Models
• Abstract summary: Applications based on large language models (LLMs) require population diversity among agents.<n>We identify a pervasive failure mode we term emphPersona Collapse: agents each assigned a distinct profile nonetheless converge into a narrow behavioral mode, producing a homogeneous simulated population.
• Score: 12.654067061944074
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Applications based on large language models (LLMs), such as multi-agent simulations, require population diversity among agents. We identify a pervasive failure mode we term \emph{Persona Collapse}: agents each assigned a distinct profile nonetheless converge into a narrow behavioral mode, producing a homogeneous simulated population. To quantify persona collapse, we propose a framework that measures how much of the persona space a population occupies (Coverage), how evenly agents spread across it (Uniformity), and how rich the resulting behavioral patterns are (Complexity). Evaluating ten LLMs on personality simulation (BFI-44), moral reasoning, and self-introduction, we observe persona collapse along two axes: (1) Dimensions: a model can appear diverse on one axis yet structurally degenerate on another, and (2) Domains: the same model may collapse the most in personality yet be the most diverse in moral reasoning. Furthermore, item-level diagnostics reveal that behavioral variation tracks coarse demographic stereotypes rather than the fine-grained individual differences specified in each persona. Counter-intuitively, \textbf{the models achieving the highest per-persona fidelity consistently produce the most stereotyped populations}. We release our toolkit and data to support population-level evaluation of LLMs.

Related papers

• Mitigating Identity Essentialism in LLM Agents with Longitudinal Life Trajectories [36.94856258384994]
  We show that static-profile agents exhibit stronger demographic separation and within-group compression than humans.<n>We propose LifeMem, a longitudinal memory framework that combines structured life-event retrieval with agent-specific parametric memory for experience integration.
  arXiv  Detail & Related papers  (2026-08-20T04:13:11Z)
• Analyzing and Correcting Benevolence Bias in Large Language Models [25.94213078450494]
  Large language models (LLMs) are increasingly used as stand-ins for human respondents.<n>Here we identify and measure benevolence bias, a small but consistent tendency for aligned LLMs to lean toward the kinder, safer, more socially approved answer on value-laden survey questions.<n>We find that the bias is a stable model property, not a quirk of any one system.
  arXiv  Detail & Related papers  (2026-07-26T08:37:58Z)
• How Well Do Large Language Models Capture Human Personality? [38.97721735513693]
  Large language models (LLMs) are increasingly used to simulate human populations via persona prompting.<n>In this work, we formalize assumptions that richer persona descriptions improve behavioral fidelity, similarly sized attribute combinations are equally simulatable, and persona definitions generalize across tasks.<n>We identify a fundamental limitation we term persona manifold collapse, where increasingly expressive persona specifications lead to systematic contraction of representational and behavioral diversity.
  arXiv  Detail & Related papers  (2026-05-12T10:33:48Z)
• Machine individuality: Separating genuine idiosyncrasy from response bias in large language models [1.4323566945483497]
  Large language models (LLMs) are increasingly integrated into daily life, in roles ranging from high-stakes decision support to companionship.<n>Here, we apply random crossed models to 74.9 million ratings provided by 10 open-weight LLMs for over 100,000 words across 14 psycholinguistic norms.<n>On average, 16.9% of variance is attributable to stimulus-specific individuality, robustly exceeding a statistical null model.
  arXiv  Detail & Related papers  (2026-04-18T00:02:41Z)
• Towards Real-world Human Behavior Simulation: Benchmarking Large Language Models on Long-horizon, Cross-scenario, Heterogeneous Behavior Traces [81.41397370235102]
  We introduce OmniBehavior, the first user simulation benchmark constructed entirely from real-world data.<n>We show that current models struggle to accurately simulate complex behaviors, with performance plateauing even as context windows expand.<n>This results in the loss of individual differences and long-tail behaviors, highlighting critical directions for future high-fidelity simulation research.
  arXiv  Detail & Related papers  (2026-04-09T15:26:21Z)
• HUMANLLM: Benchmarking and Reinforcing LLM Anthropomorphism via Human Cognitive Patterns [59.17423586203706]
  We present HUMANLLM, a framework treating psychological patterns as interacting causal forces.<n>We construct 244 patterns from 12,000 academic papers and synthesize 11,359 scenarios where 2-5 patterns reinforce, conflict, or modulate each other.<n>Our dual-level checklists evaluate both individual pattern fidelity and emergent multi-pattern dynamics, achieving strong human alignment.
  arXiv  Detail & Related papers  (2026-01-15T08:56:53Z)
• MASim: Multilingual Agent-Based Simulation for Social Science [68.04129327237963]
  Multi-agent role-playing has recently shown promise for studying social behavior with language agents.<n>Existing simulations are mostly monolingual and fail to model cross-lingual interaction.<n>We introduce MASim, the first multilingual agent-based simulation framework.
  arXiv  Detail & Related papers  (2025-12-08T06:12:48Z)
• Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond) [90.45301024940329]
  Language models (LMs) often struggle to generate diverse, human-like creative content.<n>We introduce Infinity-Chat, a large-scale dataset of 26K diverse, real-world, open-ended user queries.<n>We present a large-scale study of mode collapse in LMs, revealing a pronounced Artificial Hivemind effect.
  arXiv  Detail & Related papers  (2025-10-27T03:16:21Z)
• MF-LLM: Simulating Population Decision Dynamics via a Mean-Field Large Language Model Framework [53.82097200295448]
  Mean-Field LLM (MF-LLM) is first to incorporate mean field theory into social simulation.<n>MF-LLM models bidirectional interactions between individuals and the population through an iterative process.<n> IB-Tune is a novel fine-tuning method inspired by the Information Bottleneck principle.
  arXiv  Detail & Related papers  (2025-04-30T12:41:51Z)
• Mixture-of-Personas Language Models for Population Simulation [20.644911871150136]
  Large Language Models (LLMs) can augment human-generated data in social science research and machine learning model training.<n>MoP is a contextual mixture model, where each component is an LM agent characterized by a persona and an exemplar representing subpopulation behaviors.<n>MoP is flexible, requires no model finetuning, and is transferable across base models.
  arXiv  Detail & Related papers  (2025-04-07T12:43:05Z)
• Generative Agent Simulations of 1,000 People [56.82159813294894]
  We present a novel agent architecture that simulates the attitudes and behaviors of 1,052 real individuals. The generative agents replicate participants' responses on the General Social Survey 85% as accurately as participants replicate their own answers. Our architecture reduces accuracy biases across racial and ideological groups compared to agents given demographic descriptions.
  arXiv  Detail & Related papers  (2024-11-15T11:14:34Z)
• One fish, two fish, but not the whole sea: Alignment reduces language models' conceptual diversity [2.5975241792179378]
  Researchers have proposed using large language models (LLMs) as replacements for humans in behavioral research.<n>It is debated whether post-training alignment (RLHF or RLAIF) affects models' internal diversity.<n>We use a new way of measuring the conceptual diversity of synthetically-generated LLM "populations" by relating the internal variability of simulated individuals to the population-level variability.
  arXiv  Detail & Related papers  (2024-11-07T04:38:58Z)
• Learning signatures of decision making from many individuals playing the same game [54.33783158658077]
  We design a predictive framework that learns representations to encode an individual's 'behavioral style' We apply our method to a large-scale behavioral dataset from 1,000 humans playing a 3-armed bandit task.
  arXiv  Detail & Related papers  (2023-02-21T21:41:53Z)

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.