mlsanityData to 14 Oct 2026

CitySeeker: How Do VLMS Explore Embodied Urban Navigation With Implicit Human Needs?

arXiv 2512.16755 ↗First posted 18 Dec 2025cs.AIcited by 8

On the map: Multimodal LLMs / vision-language models → Embodied and 3D vision-language models

Abstract (arXiv)

Vision-Language Models (VLMs) have made significant progress in explicit instruction-based navigation; however, their ability to interpret implicit human needs (e.g., "I am thirsty") in dynamic urban environments remains underexplored. This paper introduces CitySeeker, a novel benchmark designed to assess VLMs' spatial reasoning and decision-making capabilities for exploring embodied urban navigation to address implicit needs. CitySeeker includes 6,440 trajectories across 8 cities, capturing diverse visual characteristics and implicit needs in 7 goal-driven scenarios. Extensive experiments reveal that even top-performing models (e.g., Qwen2.5-VL-32B-Instruct) achieve only 21.1% task completion. We find key bottlenecks in error accumulation in long-horizon reasoning, inadequate spatial cognition, and deficient experiential recall. To further analyze them, we investigate a series of exploratory strategies-Backtracking Mechanisms, Enriching Spatial Cognition, and Memory-Based Retrieval (BCR), inspired by human cognitive mapping's emphasis on iterative observation-reasoning cycles and adaptive path optimization. Our analysis provides actionable insights for developing VLMs with robust spatial intelligence required for tackling "last-mile" navigation challenges.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 47 of its 53 references that are in the map. First posted on arXiv 18 Dec 2025.

Builds on

47 of 53 references.

Closest ancestors · 7

Same direction, newest first

  1. 2025MindCube: Spatial Mental Modeling from Limited Views26 Jun
  2. 2025Praxis-VLM: Vision-Grounded Decision Making via Text-Driven Reinforcement Learning21 Mar
  3. 2024Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces18 Dec
  4. 2024End-to-End Navigation with Vision Language Models: Transforming Spatial Reasoning into Question-Answering8 Nov
  5. 2024Robot Navigation Using Physically Grounded Vision-Language Models in Outdoor Environments30 Sep
  6. 2024FLAME: Learning to Navigate with Multimodal LLM in Urban Environments20 Aug
  7. 2024NavGPT-2: Unleashing Navigational Reasoning Capability for Large Vision-Language Models17 Jul

Other references · 40

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  3. 2023Visual Instruction Tuningcited by 7,869
  4. 2025Qwen2.5-VL Technical Reportcited by 5,587
  5. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  6. 2023InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Taskscited by 2,565
  7. 2024Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phonecited by 1,917
  8. 2024MiniCPM-V: A GPT-4V Level MLLM on Your Phonecited by 1,079
  9. 2017Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environmentscited by 1,027
  10. 2023Vision-Language Models for Vision Tasks: A Surveycited by 839
  11. 2025Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAscited by 521
  12. 2019REVERIE: Remote Embodied Visual Referring Expression in Real Indoor Environmentscited by 333
28 more
  1. 2023Language Models Represent Space and Timecited by 316
  2. 2023NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large Language Modelscited by 280
  3. 2018Touchdown: Natural Language Navigation and Spatial Reasoning in Visual Street Environmentscited by 239
  4. 2022ZSON: Zero-Shot Object-Goal Navigation using Multimodal Goal Embeddingscited by 178
  5. 2025MiniMax-01: Scaling Foundation Models with Lightning Attentioncited by 166
  6. 2024RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthinesscited by 131
  7. 2018Learning to Navigate in Cities Without a Mapcited by 104
  8. 2021Structured Scene Memory for Vision-Language Navigationcited by 95
  9. 2024Vision-and-Language Navigation Today and Tomorrow: A Survey in the Era of Foundation Modelscited by 79
  10. 2023Evaluating Cognitive Maps and Planning in Large Language Models with CogEvalcited by 77
  11. 2024Mobility VLA: Multimodal Instruction Navigation with Long-Context VLMs and Topological Graphscited by 73
  12. 2023VELMA: Verbalization Embodiment of LLM Agents for Vision and Language Navigation in Street Viewcited by 67
  13. 2024Mind's Eye of LLMs: Visualization-of-Thought Elicits Spatial Reasoning in Large Language Modelscited by 66
  14. 2023SACSoN: Scalable Autonomous Control for Social Navigationcited by 61
  15. 2022CLIP-Nav: Using CLIP for Zero-Shot Vision-and-Language Navigationcited by 55
  16. 2023LangNav: Language as a Perceptual Representation for Navigationcited by 47
  17. 2018Talk the Walk: Navigating New York City through Grounded Dialoguecited by 44
  18. 2024CityWalker: Learning Embodied Urban Navigation from Web-Scale Videoscited by 41
  19. 2021Vision-Language Navigation: A Survey and Taxonomycited by 34
  20. 2024CogNav: Cognitive Process Modeling for Object Goal Navigation with LLMscited by 33
  21. 2024CityGPT: Empowering Urban Spatial Cognition of Large Language Modelscited by 30
  22. 2024MC-GPT: Empowering Vision-and-Language Navigation with Memory Map and Reasoning Chainscited by 25
  23. 2020Generating Landmark Navigation Instructions from Maps as a Graph-to-Text Problemcited by 22
  24. 2023Find What You Want: Learning Demand-conditioned Object Attribute Space for Demand-driven Navigationcited by 18
  25. 2024Vision-Language Navigation with Embodied Intelligence: A Surveycited by 9
  26. 2023StreetNav: Leveraging Street Cameras to Support Precise Outdoor Navigation for Blind Pedestrianscited by 7
  27. 2025Mind Meets Space: Rethinking Agentic Spatial Intelligence from a Neuroscience-inspired Perspectivecited by 7
  28. 2024Decoding Urban Industrial Complexity: Enhancing Knowledge-Driven Insights via IndustryScopeGPTcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 8.

8 papers.

  1. 2026MineExplorer: Evaluating Open-World Exploration of MLLM Agents in Minecraftcited by 4
  2. 2026POINav: Benchmarking and Enhancing Final-Meters Arrival in Real-World Vision-Language Navigationcited by 2
  3. 2026DeliveryGym: An RL Environment for Long-Horizon Embodied Agent Planning with Adaptive Curriculumcited by 2
  4. 2026SpatialFly: Implicit 3D Prior-Guided Visual Reparameterization for Continuous UAV Vision-and-Language Navigationcited by 1
  5. 2026360CityArena: A Realistic Virtual Urban Navigation Benchmark for Embodied Agentscited by 1
  6. 2026UrbanGround: From Local Perception to Spatial Agency in a Real-Scale Citycited by 1
  7. 2026Towards Embodied Air-Ground Cooperative Object Search: Benchmark, Dataset and Agentic Methodcited by 1
  8. 2026MAP: A Benchmark on Multimodal Accessibility Planning for Real World Placescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025City Navigation in the Wild: Exploring Emergent Navigation from Web-Scale Knowledge in MLLMs
  2. 2025How Far are VLMs from Visual Spatial Intelligence? A Benchmark-Driven Perspective
  3. 2026What do VLM-Based Vision-Language Navigation Models Rely on: Interpreting and Steering Policy Behavior
  4. 2025UrbanVideo-Bench: Benchmarking Vision-Language Models on Embodied Intelligence with Video Data in Urban Spaces
  5. 2025A Survey of Large Language Model-Powered Spatial Intelligence Across Scales: Advances in Embodied Agents, Smart Cities, and Earth Science
  6. 2026CityCube: Benchmarking Cross-view Spatial Reasoning on Vision-Language Models in Urban Environments
  7. 2025Mem4Nav: Boosting Vision-and-Language Navigation in Urban Environments with a Hierarchical Spatial-Cognition Long-Short Memory System
  8. 2026VCN-Bench: A Video-Contextualized Navigation Benchmark for Spatial Reasoning over Prior Visual Experience