mlsanityData to 7 Oct 2026

Foundation Models for Geospatial Reasoning: Assessing Capabilities of Large Language Models in Understanding Geometries and Topological Spatial Relations

arXiv 2505.17136 ↗First posted 22 May 2025cs.CL cs.AIcited by 20

On the map: LLM reasoning → Spatial and embodied reasoning

Abstract (arXiv)

Applying AI foundation models directly to geospatial datasets remains challenging due to their limited ability to represent and reason with geographical entities, specifically vector-based geometries and natural language descriptions of complex spatial relations. To address these issues, we investigate the extent to which a well-known-text (WKT) representation of geometries and their spatial relations (e.g., topological predicates) are preserved during spatial reasoning when the geospatial vector data are passed to large language models (LLMs) including GPT-3.5-turbo, GPT-4, and DeepSeek-R1-14B. Our workflow employs three distinct approaches to complete the spatial reasoning tasks for comparison, i.e., geometry embedding-based, prompt engineering-based, and everyday language-based evaluation. Our experiment results demonstrate that both the embedding-based and prompt engineering-based approaches to geospatial question-answering tasks with GPT models can achieve an accuracy of over 0.6 on average for the identification of topological spatial relations between two geometries. Among the evaluated models, GPT-4 with few-shot prompting achieved the highest performance with over 0.66 accuracy on topological spatial relation inference. Additionally, GPT-based reasoner is capable of properly comprehending inverse topological spatial relations and including an LLM-generated geometry can enhance the effectiveness for geographic entity retrieval. GPT-4 also exhibits the ability to translate certain vernacular descriptions about places into formal topological relations, and adding the geometry-type or place-type context in prompts may improve inference accuracy, but it varies by instance. The performance of these spatial reasoning tasks offers valuable insights for the refinement of LLMs with geographical knowledge towards the development of geo-foundation models capable of geospatial reasoning.

Key facts. Cited by 20 papers in the map (data through 7 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 41 of its 113 references that are in the map. First posted on arXiv 22 May 2025.

Builds on

41 of 113 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces18 Dec
  2. 2024Can Large Language Models Reason about the Region Connection Calculus?29 Nov
  3. 2023An Evaluation of ChatGPT-4's Qualitative Spatial Reasoning Capabilities in RCC-827 Sep
  4. 2023Dialectical language model evaluation: An initial appraisal of the commonsense spatial reasoning abilities of LLMs22 Apr

Other references · 37

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2020Language Models are Few-Shot Learnerscited by 32,199
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  4. 2023GPT-4 Technical Reportcited by 18,132
  5. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  6. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  7. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,049
  8. 2023Segment Anythingcited by 8,158
  9. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  10. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  11. 2022Flamingo: a Visual Language Model for Few-Shot Learningcited by 4,337
  12. 2022Large Language Models are Zero-Shot Reasonerscited by 4,323
25 more
  1. 2022LaMDA: Language Models for Dialog Applicationscited by 1,218
  2. 2022Text and Code Embeddings by Contrastive Pre-Trainingcited by 358
  3. 2022GatorTron: A Large Clinical Language Model to Unlock Patient Information from Unstructured Electronic Health Recordscited by 238
  4. 2022SGPT: GPT Sentence Embeddings for Semantic Searchcited by 204
  5. 2018An efficient framework for learning sentence representationscited by 197
  6. 2023Query2doc: Query Expansion with Large Language Modelscited by 146
  7. 2023Foundation Models for Generalist Geospatial Artificial Intelligencecited by 143
  8. 2023Evaluating Spatial Understanding of Large Language Modelscited by 65
  9. 2023GeoLLM: Extracting Geospatial Knowledge from Large Language Modelscited by 62
  10. 2021A Review of Location Encoding for GeoAI: Methods and Applicationscited by 53
  11. 2021Adversarial Text-to-Image Synthesis: A Reviewcited by 52
  12. 2023Autonomous GIS: the next-generation AI-powered GIScited by 52
  13. 2023Geo-knowledge-guided GPT models improve the extraction of location descriptions from disaster-related social media messagescited by 31
  14. 2024Towards Reproducible LLM Evaluation: Quantifying Uncertainty in LLM Benchmark Scorescited by 25
  15. 2020SE-KGE: A Location-Aware Knowledge Graph Embedding Model for Geographic Question Answering and Spatial Semantic Liftingcited by 23
  16. 2023Assessment of a new GeoAI foundation model for flood inundation mappingcited by 14
  17. 2023Building Privacy-Preserving and Secure Geospatial Artificial Intelligence Foundation Modelscited by 12
  18. 2023Philosophical Foundations of GeoAI: Exploring Sustainability, Diversity, and Bias in GeoAI and Spatial Data Sciencecited by 11
  19. 2024Evaluating the Ability of Large Language Models to Reason about Cardinal Directionscited by 10
  20. 2023Evaluating the Effectiveness of Large Language Models in Representing Textual Descriptions of Geometry and Spatial Relationscited by 9
  21. 2024Evaluation of Code LLMs on Geospatial Code Generationcited by 8
  22. 2024A systematic review of geospatial location embedding approaches in large language models: A path to spatial AI systemscited by 6
  23. 2024Distortions in Judged Spatial Relations in Large Language Modelscited by 5
  24. 2023Core Building Blocks: Next Gen Geo Spatial GPT Applicationcited by 4
  25. 2022DeepSSN: a deep convolutional neural network to assess spatial scene similaritycited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 7; 2026: 13.

20 papers.

Sort:

By area:

  1. 2025GeoAnalystBench: A GeoAI benchmark for assessing large language models for spatial analysis workflow and code generationcited by 9
  2. 2025A Vision for Geo-Temporal Deep Research Systems: Towards Comprehensive, Transparent, and Reproducible Geo-Temporal Information Synthesiscited by 2
  3. 2026GeoAgentBench: A Dynamic Execution Benchmark for Tool-Augmented Agents in Spatial Analysiscited by 2
  4. 2025GeoResponder: Towards Building Geospatial LLMs for Time-Critical Disaster Responsecited by 1
  5. 2025Georeferencing complex relative locality descriptions with large language modelscited by 1
  6. 2025Thinking on Maps: How Foundation Model Agents Explore, Remember, and Reason Map Environmentscited by 1
  7. 2026Mil-SCORE: Benchmarking Long-Context Geospatial Reasoning and Planning in Large Language Modelscited by 1
  8. 2026A Dual-Helix Governance Approach Towards Reliable Agentic Artificial Intelligence for WebGIS Developmentcited by 1
  9. 2025From Pixels to Urban Policy-Intelligence: Recovering Legacy Effects of Redlining with a Multimodal LLMcited by 0
  10. 2025Coordinates from Context: Using LLMs to Ground Complex Location Referencescited by 0
  11. 2026Geography According to ChatGPT -- How Generative AI Represents and Reasons about Geographycited by 0
  12. 2026Unbox Responsible GeoAI: Navigating Climate Extreme and Disaster Mappingcited by 0
8 more
  1. 2026Spatial Representation Learning Beyond Pixels: Unifying Raster Data and Vector Semantics for Human-Centric Geospatial Foundation Modelscited by 0
  2. 2026Assessing the Geographic Diversity of AI's Platial Representations in Image Generationcited by 0
  3. 2026Multimodal and Multiscale Spatial-Temporal Semantic Search and Recommendation with AI Foundation Modelscited by 0
  4. 2026Do Maps Still Matter for Machines: Revisiting the Role of Choropleth Maps in Foundation Model Spatial Understandingcited by 0
  5. 2026LMM Modality Transfer: A Pre-requisite for Autonomous GIS Agentscited by 0
  6. 2026Geo-Spatial Concept Probing of Large Language Models: Abstraction, Compositionality, and Groundingcited by 0
  7. 2026Physico-Geospatial Grounded Scene Interpretation for Mobile Roboticscited by 0
  8. 2026PolyTopoBench: A Benchmark for Complex Vector Polygon Generation from Remote Sensing Imagerycited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026MultiGlobeQA: A Multilingual and Globally Diverse Benchmark for Geospatial Reasoning
  2. 2026From Symbolic to Geometric: Enabling Spatial Reasoning in Large Language Models
  3. 2024SpaRC and SpaRP: Spatial Reasoning Characterization and Path Generation for Understanding Spatial Reasoning Capability of Large Language Models
  4. 2026GPSBench: Do Large Language Models Understand GPS Coordinates?
  5. 2024MapEval: A Map-Based Evaluation of Geo-Spatial Reasoning in Foundation Models
  6. 2026MapVerse: A Benchmark for Geospatial Question Answering on Diverse Real-World Maps
  7. 2025Stuck in the Matrix: Probing Spatial Reasoning in Large Language Models
  8. 2025Can Large Language Models Integrate Spatial Data? Empirical Insights into Reasoning Strengths and Computational Weaknesses