arXiv:2510.04391v3 Announce Type: replace-cross 
Abstract: The computational role of imagination remains debated. While classical accounts emphasize reward maximization, emerging evidence suggests imagination serves a broader function: accessing internal world models (IWMs). Here, we employ psychological network analysis to compare IWMs in humans and large language models (LLMs) through imagination vividness ratings. Using the Vividness of Visual Imagery Questionnaire (VVIQ-2) and Plymouth Sensory Imagery Questionnaire (PSIQ), we construct imagination networks from three human populations (Florida, Poland, London; N=2,743) and six LLM variants in two conversation conditions. Human imagination networks demonstrate robust correlations across centrality measures (expected influence, strength, closeness) and consistent clustering patterns, indicating shared structural organization of IWMs across populations. In contrast, LLM-derived networks show minimal clustering and weak centrality correlations, even when manipulating conversational memory. These systematic differences persist across environmental scenes (VVIQ-2) and sensory modalities (PSIQ), revealing fundamental disparities between human and artificial world models. Our network-based approach provides a quantitative framework for comparing internally-generated representations across cognitive agents, with implications for developing human-like imagination in artificial intelligence systems.

دراسة حديثة نُشرت على arXiv تستكشف دور الخيال في الوكلاء المعرفيين، من خلال مقارنة النماذج الداخلية للعالم (IWMs) لدى البشر ونماذج اللغة الكبيرة (LLMs). باستخدام استبيان حيوية التصور البصري واستبيان التصور الحسي في بليموث، شملت الدراسة 2743 مشاركًا من فلوريدا وبولندا ولندن، وكشفت عن ارتباطات قوية في شبكات الخيال البشرية، بينما أظهرت الشبكات المستمدة من LLM تكتلًا ضئيلًا وارتباطات مركزية ضعيفة.

Un estudio reciente publicado en arXiv explora el papel de la imaginación en los agentes cognitivos, comparando los modelos internos del mundo (IWMs) en humanos y modelos de lenguaje grandes (LLMs). Utilizando el Cuestionario de Vividness of Visual Imagery y el Cuestionario de Imaginación Sensorial de Plymouth, la investigación involucró a 2,743 participantes de Florida, Polonia y Londres, revelando correlaciones robustas en las redes de imaginación humanas, mientras que las redes de LLM mostraron un agrupamiento mínimo y correlaciones de centralidad débiles.

Une étude récente publiée sur arXiv examine le rôle de l'imagination dans les agents cognitifs, en comparant les modèles internes du monde (IWMs) chez les humains et les grands modèles de langage (LLMs). En utilisant le Vividness of Visual Imagery Questionnaire et le Plymouth Sensory Imagery Questionnaire, la recherche a impliqué 2 743 participants de Floride, de Pologne et de Londres, révélant des corrélations robustes dans les réseaux d'imagination humains, tandis que les réseaux LLM présentaient un faible regroupement et de faibles corrélations de centralité.

A recent study published on arXiv explores the role of imagination in cognitive agents, comparing internal world models (IWMs) in humans and large language models (LLMs). Utilizing the Vividness of Visual Imagery Questionnaire and Plymouth Sensory Imagery Questionnaire, the research involved 2,743 participants from Florida, Poland, and London, revealing robust correlations in human imagination networks, while LLM networks exhibited minimal clustering and weak centrality correlations.

Internal World Models as Imagination Networks in Cognitive Agents

Was this article worth reading? Share it

LucidQuery AI

Magicley AI

SwapAnything.io