mlsanityData to 14 Oct 2026

GeoChain: Multimodal Chain-of-Thought for Geographic Reasoning

arXiv 2506.00785 ↗First posted 1 Jun 2025cs.AI cs.CV cs.LGcited by 16

On the map: LLM reasoning → Spatial and embodied reasoning

also in 1 more
Abstract (arXiv)

This paper introduces GeoChain, a large-scale benchmark for evaluating step-by-step geographic reasoning in multimodal large language models (MLLMs). Leveraging 1.46 million Mapillary street-level images, GeoChain pairs each image with a 21-step chain-of-thought (CoT) question sequence (over 30 million Q&A pairs). These sequences guide models from coarse attributes to fine-grained localization across four reasoning categories - visual, spatial, cultural, and precise geolocation - annotated by difficulty. Images are also enriched with semantic segmentation (150 classes) and a visual locatability score. Our benchmarking of contemporary MLLMs (GPT-4.1 variants, Claude 3.7, Gemini 2.5 variants) on a diverse 2,088-image subset reveals consistent challenges: models frequently exhibit weaknesses in visual grounding, display erratic reasoning, and struggle to achieve accurate localization, especially as the reasoning complexity escalates. GeoChain offers a robust diagnostic methodology, critical for fostering significant advancements in complex geographic reasoning within MLLMs.

Key facts. Cited by 16 papers in the map (data through 14 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 14 of its 33 references that are in the map. First posted on arXiv 1 Jun 2025.

Builds on

14 of 33 references.

References · 14

  1. 2023GPT-4 Technical Reportcited by 18,396
  2. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  3. 2024Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextcited by 3,224
  4. 2023InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuningcited by 2,876
  5. 2025Humanity's Last Examcited by 491
  6. 2023GEO-Bench: Toward Foundation Models for Earth Monitoringcited by 124
  7. 2025Geolocation with Real Human Gameplay Data: A Large-Scale Dataset and Human-Like Reasoning Frameworkcited by 24
  8. 2025AI Sees Your Location, But With A Bias Toward The Wealthy Worldcited by 14
  9. 2025Evaluating Precise Geolocation Inference Capabilities of Vision Language Modelscited by 12
  10. 2023MAEA: Multimodal Attribution for Embodied AIcited by 9
  11. 2025HueManity: Probing Fine-Grained Visual Perception in MLLMscited by 9
  12. 2024Semantic Augmentation in Images using Languagecited by 7
2 more
  1. 2024Attribution Regularization for Multimodal Paradigmscited by 7
  2. 2025GAEA: A Geolocation Aware Conversational Assistantcited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 10; 2026: 6.

16 papers.

Sort:

By area:

  1. 2025CountQA: How Well Do MLLMs Count in the Wild?cited by 19
  2. 2025Towards Faithful Reasoning in Remote Sensing: A Perceptually-Grounded GeoSpatial Chain-of-Thought for Vision-Language Modelscited by 19
  3. 2025GeoArena: Evaluating Open-World Geographic Reasoning in Large Vision-Language Modelscited by 5
  4. 2025MaRVL-QA: A Benchmark for Mathematical Reasoning over Visual Landscapescited by 4
  5. 2025The Artificial Intelligence Cognitive Examination: A Survey on the Evolution of Multimodal Evaluation from Recognition to Reasoningcited by 4
  6. 2025Explicit Reasoning Makes Better Judges: A Systematic Study on Accuracy, Efficiency, and Robustnesscited by 3
  7. 2025Unlocking Zero-Shot Geospatial Reasoning via Indirect Rewardscited by 2
  8. 2025GEO-Detective: Unveiling Location Privacy Risks in Images with LLM Agentscited by 1
  9. 2026Uni-RS: A Spatially Faithful Unified Understanding and Generation Model for Remote Sensingcited by 1
  10. 2026LocationAgent: A Hierarchical Agent for Image Geolocation via Decoupling Strategy and Evidence from Parametric Knowledgecited by 1
  11. 2026UrbanGraphEmbeddings: Learning and Evaluating Spatially Grounded Multimodal Embeddings for Urban Sciencecited by 1
  12. 2026ENC-Bench: A Benchmark for Evaluating Multimodal Large Language Models in Electronic Navigational Chart Understandingcited by 1
4 more
  1. 2025Confidence Aware SSD Ensemble with Weighted Boxes Fusion for Weapon Detectioncited by 0
  2. 2025Understanding Virality: A Rubric based Vision-Language Model Framework for Short-Form Edutainment Evaluationcited by 0
  3. 2026GTPred: Benchmarking MLLMs for Interpretable Geo-localization and Time-of-capture Predictioncited by 0
  4. 2026GeoRouter: Dynamic Paradigm Routing for Worldwide Image Geolocalizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Where on Earth? A Vision-Language Benchmark for Probing Model Geolocation Skills Across Scales
  2. 2026Mind the Spatial Gap: A Benchmark for Geospatial Reasoning and Constrained Spatial Editing
  3. 2026MapVerse: A Benchmark for Geospatial Question Answering on Diverse Real-World Maps
  4. 2025Recognition through Reasoning: Reinforcing Image Geo-localization with Large Vision-Language Models
  5. 2026GeoRC: A Benchmark for Geolocation Reasoning Chains
  6. 2025GRE Suite: Geo-localization Inference via Fine-Tuned Vision-Language Models and Enhanced Reasoning Chains
  7. 2026Where Do Vision-Language Models Fail? World Scale Analysis for Image Geolocalization
  8. 2026SpatialChain: A Benchmark for Auditing Spatial Reasoning Faithfulness in VLMs