# Evaluating Spatial Understanding of Large Language Models

arXiv: https://arxiv.org/abs/2310.14540 · HTML: https://mlsanity.com/paper/2310.14540/

## Key facts

- Cited by 66 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 21 references that are in the map.
- First posted on arXiv 23 Oct 2023.

## Builds on (most cited first, 10 of 10)

- [Sparks of Artificial General Intelligence: Early experiments with GPT-4](https://mlsanity.com/paper/2303.12712/) (2023)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [Code as Policies: Language Model Programs for Embodied Control](https://mlsanity.com/paper/2209.07753/) (2022)
- [Inner Monologue: Embodied Reasoning through Planning with Language Models](https://mlsanity.com/paper/2207.05608/) (2022)
- [Implicit Representations of Meaning in Neural Language Models](https://mlsanity.com/paper/2106.00737/) (2021)
- [Can Language Models Encode Perceptual Structure Without Grounding? A Case Study in Color](https://mlsanity.com/paper/2109.06129/) (2021)
- [SpartQA: : A Textual Question Answering Benchmark for Spatial Reasoning](https://mlsanity.com/paper/2104.05832/) (2021)
- [Evaluating Cognitive Maps and Planning in Large Language Models with CogEval](https://mlsanity.com/paper/2309.15129/) (2023)
- [StepGame: A New Benchmark for Robust Multi-Hop Spatial Reasoning in Texts](https://mlsanity.com/paper/2204.08292/) (2022)
- [SayNav: Grounding Large Language Models for Dynamic Planning to Navigation in New Environments](https://mlsanity.com/paper/2309.04077/) (2023)

## Built on by (most cited first, 30 of 66)

- [Gemma 3 Technical Report](https://mlsanity.com/paper/2503.19786/) (2025)
- [Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces](https://mlsanity.com/paper/2412.14171/) (2024)
- [BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games](https://mlsanity.com/paper/2411.13543/) (2024)
- [RoboSpatial: Teaching Spatial Understanding to 2D and 3D Vision-Language Models for Robotics](https://mlsanity.com/paper/2411.16537/) (2024)
- [Reinforcing Spatial Reasoning in Vision-Language Models with Interwoven Thinking and Visual Drawing](https://mlsanity.com/paper/2506.09965/) (2025)
- [BIG-Bench Extra Hard](https://mlsanity.com/paper/2502.19187/) (2025)
- [Mind's Eye of LLMs: Visualization-of-Thought Elicits Spatial Reasoning in Large Language Models](https://mlsanity.com/paper/2404.03622/) (2024)
- [SpatialBench: Benchmarking Multimodal Large Language Models for Spatial Cognition](https://mlsanity.com/paper/2511.21471/) (2025)
- [Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMs](https://mlsanity.com/paper/2507.02076/) (2025)
- [CityGPT: Empowering Urban Spatial Cognition of Large Language Models](https://mlsanity.com/paper/2406.13948/) (2024)
- [VideoINSTA: Zero-shot Long Video Understanding via Informative Spatial-Temporal Reasoning with LLMs](https://mlsanity.com/paper/2409.20365/) (2024)
- [iVISPAR -- An Interactive Visual-Spatial Reasoning Benchmark for VLMs](https://mlsanity.com/paper/2502.03214/) (2025)
- [Foundation Models for Geospatial Reasoning: Assessing Capabilities of Large Language Models in Understanding Geometries and Topological Spatial Relations](https://mlsanity.com/paper/2505.17136/) (2025)
- [Whiteboard-of-Thought: Thinking Step-by-Step Across Modalities](https://mlsanity.com/paper/2406.14562/) (2024)
- [SpatialThinker: Reinforcing Scene Graph-Grounded Spatial Reasoning via Dense Rewards](https://mlsanity.com/paper/2511.07403/) (2025)
- [FloorplanQA: A Benchmark for Spatial Reasoning in LLMs using Structured Representations](https://mlsanity.com/paper/2507.07644/) (2025)
- [Scaffolding Language Learning via Multi-modal Tutoring Systems with Pedagogical Instructions](https://mlsanity.com/paper/2404.03429/) (2024)
- [Evaluating Large Language Models on Spatial Tasks: A Multi-Task Benchmarking Study](https://mlsanity.com/paper/2408.14438/) (2024)
- [A Survey of Large Language Model-Powered Spatial Intelligence Across Scales: Advances in Embodied Agents, Smart Cities, and Earth Science](https://mlsanity.com/paper/2504.09848/) (2025)
- [STBench: Assessing the Ability of Large Language Models in Spatio-Temporal Analysis](https://mlsanity.com/paper/2406.19065/) (2024)
- [Evaluating the Ability of Large Language Models to Reason about Cardinal Directions](https://mlsanity.com/paper/2406.16528/) (2024)
- [From Text to Space: Mapping Abstract Spatial Models in LLMs during a Grid-World Navigation Task](https://mlsanity.com/paper/2502.16690/) (2025)
- [El Agente Quntur: A research collaborator agent for quantum chemistry](https://mlsanity.com/paper/2602.04850/) (2026)
- [How language models extrapolate outside the training data: A case study in Textualized Gridworld](https://mlsanity.com/paper/2406.15275/) (2024)
- [GRASP: A Grid-Based Benchmark for Evaluating Commonsense Spatial Reasoning](https://mlsanity.com/paper/2407.01892/) (2024)
- [Grammar-based Game Description Generation using Large Language Models](https://mlsanity.com/paper/2407.17404/) (2024)
- [Factorio Learning Environment](https://mlsanity.com/paper/2503.09617/) (2025)
- [Evaluating Large Language Models for Real-World Engineering Tasks](https://mlsanity.com/paper/2505.13484/) (2025)
- [Evaluating the Generation of Spatial Relations in Text and Image Generative Models](https://mlsanity.com/paper/2411.07664/) (2024)
- [Evaluating Vision-Language Models as Evaluators in Path Planning](https://mlsanity.com/paper/2411.18711/) (2024)

## Abstract (arXiv)

Large language models (LLMs) show remarkable capabilities across a variety of tasks. Despite the models only seeing text in training, several recent studies suggest that LLM representations implicitly capture aspects of the underlying grounded concepts. Here, we explore LLM representations of a particularly salient kind of grounded knowledge -- spatial relationships. We design natural-language navigation tasks and evaluate the ability of LLMs, in particular GPT-3.5-turbo, GPT-4, and Llama2 series models, to represent and reason about spatial structures. These tasks reveal substantial variability in LLM performance across different spatial structures, including square, hexagonal, and triangular grids, rings, and trees. In extensive error analysis, we find that LLMs' mistakes reflect both spatial and non-spatial factors. These findings suggest that LLMs appear to capture certain aspects of spatial structure implicitly, but room for improvement remains.