Do LLMs Build Spatial World Models? Evidence from Grid-World Maze Tasks

• URL: http://arxiv.org/abs/2604.10690v1
• Date: Sun, 12 Apr 2026 15:30:43 GMT
• Title: Do LLMs Build Spatial World Models? Evidence from Grid-World Maze Tasks
• Abstract summary: We systematically evaluate the spatial understanding of large language models through maze tasks.<n>We uncover significant discrepancies in spatial reasoning that challenge assumptions about planning capabilities.<n>These results have critical implications for deploying foundation models in applications requiring spatial abstraction.
• Score: 4.8901522421148735
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Foundation models have shown remarkable performance across diverse tasks, yet their ability to construct internal spatial world models for reasoning and planning remains unclear. We systematically evaluate the spatial understanding of large language models through maze tasks, a controlled testing context requiring multi-step planning and spatial abstraction. Across comprehensive experiments with Gemini-2.5-Flash, GPT-5-mini, Claude-Haiku-4.5, and DeepSeek-Chat, we uncover significant discrepancies in spatial reasoning that challenge assumptions about LLM planning capabilities. Using chain-of-thought prompting, Gemini achieves 80-86% accuracy on smaller mazes (5x5 to 7x7 grids) with tokenized adjacency representations, but performance collapses to 16-34% with visual grid formats, which is a 2-5x difference, suggesting representation-dependent rather than format-invariant spatial reasoning. We further probe spatial understanding through sequential proximity questions and compositional distance comparisons. Despite achieving 96-99% semantic coverage in reasoning traces, models fail to leverage this understanding for consistent spatial computations, indicating that they treat each question independently rather than building cumulative spatial knowledge. Our findings based on the maze-solving tasks suggest that LLMs do not develop robust spatial world models, but rather exhibit representation-specific and prompting-dependent reasoning that succeeds only under narrow conditions. These results have critical implications for deploying foundation models in applications requiring spatial abstraction.

Related papers

• SpatialImaginer: Towards Adaptive Visual Imagination for Spatial Reasoning [67.67774742200626]
  spatial intelligence, which refers to the ability to reason about geometric and physical structure from visual observations, is a core challenge for large language models.<n>We propose a unified multimodal generation framework that integrates textual reasoning with visual imagination.<n>Our framework adopts a divide-and-conquer strategy, using text chain-of-thought for high-level semantic planning and the visual imagination for geometry-sensitive state transformation and consistency preservation.
  arXiv  Detail & Related papers  (2026-04-19T11:21:59Z)
• From Indoor to Open World: Revealing the Spatial Reasoning Gap in MLLMs [65.04549036809557]
  We introduce a benchmark built from pedestrian-perspective videos captured with synchronized stereo cameras, LiDAR, and IMU/GPS sensors.<n>This dataset provides metrically precise 3D information, enabling the automatic generation of spatial reasoning questions.<n> Evaluations reveal that the performance gains observed in structured indoor benchmarks vanish in open-world settings.
  arXiv  Detail & Related papers  (2025-12-22T18:58:12Z)
• SpatialThinker: Reinforcing 3D Reasoning in Multimodal LLMs via Spatial Rewards [37.39035418889281]
  We introduce SpatialThinker, a 3D-aware MLLM trained with RL to integrate structured spatial grounding with multi-step reasoning.<n>The model simulates human-like spatial perception by constructing a scene graph of task-relevant objects and spatial relations, and reasoning towards an answer via dense spatial rewards.
  arXiv  Detail & Related papers  (2025-11-10T18:52:47Z)
• Multimodal Spatial Reasoning in the Large Model Era: A Survey and Benchmarks [108.15756345836901]
  We provide a comprehensive review of multimodal spatial reasoning tasks with large models.<n>We review advances in embodied AI, including vision-language navigation and action models.<n>We consider emerging modalities such as audio and egocentric video, which contribute to novel spatial understanding through new sensors.
  arXiv  Detail & Related papers  (2025-10-29T17:55:43Z)
• Stuck in the Matrix: Probing Spatial Reasoning in Large Language Models [0.0]
  This paper explores the spatial reasoning capability of large language models (LLMs) over textual input through a suite of five tasks.<n>The models were tested on both fundamental spatial reasoning and multi-step problem-solving within structured grid-based environments.
  arXiv  Detail & Related papers  (2025-10-23T04:32:46Z)
• FloorplanQA: A Benchmark for Spatial Reasoning in LLMs using Structured Representations [78.65988445433844]
  FloorplanQA is a diagnostic benchmark for evaluating spatial reasoning in large-language models.<n>The benchmark covers core spatial tasks, including distance measurement, visibility, path finding, and object placement within constrained spaces.
  arXiv  Detail & Related papers  (2025-07-10T11:16:48Z)
• SpatialScore: Towards Unified Evaluation for Multimodal Spatial Understanding [64.15606979785355]
  Multimodal large language models (MLLMs) have achieved impressive success in question-answering tasks, yet their capabilities for spatial understanding are less explored.<n>This work investigates a critical question: do existing MLLMs possess 3D spatial perception and understanding abilities?
  arXiv  Detail & Related papers  (2025-05-22T17:59:03Z)
• EmbodiedVSR: Dynamic Scene Graph-Guided Chain-of-Thought Reasoning for Visual Spatial Tasks [24.41705039390567]
  EmbodiedVSR (Embodied Visual Spatial Reasoning) is a novel framework that integrates dynamic scene graph-guided Chain-of-Thought (CoT) reasoning.<n>Our method enables zero-shot spatial reasoning without task-specific fine-tuning.<n>Experiments demonstrate that our framework significantly outperforms existing MLLM-based methods in accuracy and reasoning coherence.
  arXiv  Detail & Related papers  (2025-03-14T05:06:07Z)
• Sparkle: Mastering Basic Spatial Capabilities in Vision Language Models Elicits Generalization to Spatial Reasoning [36.588008658084895]
  Vision language models (VLMs) perform well on many tasks but often fail at spatial reasoning.<n>Our evaluation shows that state-of-the-art VLMs give implausible or incorrect answers to composite spatial problems.<n>We enhance 2D spatial reasoning in VLMs by training them only on basic spatial capabilities.
  arXiv  Detail & Related papers  (2024-10-21T16:26:09Z)
• SpaRC and SpaRP: Spatial Reasoning Characterization and Path Generation for Understanding Spatial Reasoning Capability of Large Language Models [70.01883340129204]
  spatial reasoning is a crucial component of both biological and artificial intelligence. We present a comprehensive study of the capability of current state-of-the-art large language models (LLMs) on spatial reasoning.
  arXiv  Detail & Related papers  (2024-06-07T01:06:34Z)

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.