mlsanityData to 7 Oct 2026

Think with 3D: Geometric Imagination Grounded Spatial Reasoning from Limited Views

arXiv 2510.18632 ↗First posted 21 Oct 2025cs.CV cs.AIcited by 58

On the map: LLM reasoning → Spatial and embodied reasoning

also in 1 more
Abstract (arXiv)

Though recent advances in vision-language models (VLMs) have achieved remarkable progress across a wide range of multimodal tasks, understanding 3D spatial relationships from limited views remains a significant challenge. Previous reasoning methods typically rely on pure text (e.g., topological cognitive maps) or on 2D visual cues. However, their limited representational capacity hinders performance in specific tasks that require 3D spatial imagination. To address this limitation, we propose 3DThinker, a framework that can effectively exploits the rich geometric information embedded within images while reasoning, like humans do. Our framework is the first to enable 3D mentaling during reasoning without any 3D prior input, and it does not rely on explicitly labeled 3D data for training. Specifically, our training consists of two stages. First, we perform supervised training to align the 3D latent generated by VLM while reasoning with that of a 3D foundation model (e.g., VGGT). Then, we optimize the entire reasoning trajectory solely based on outcome signals, thereby refining the underlying 3D mentaling. Extensive experiments across multiple benchmarks show that 3DThinker consistently outperforms strong baselines and offers a new perspective toward unifying 3D representations into multimodal reasoning. Our code is available at https://github.com/zhangquanchen/3DThinker.

Key facts. Cited by 58 papers in the map (data through 7 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 66 of its 98 references that are in the map. First posted on arXiv 21 Oct 2025.

Builds on

66 of 98 references.

Closest ancestors · 15

Same direction, newest first

  1. 2025SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language Models9 Oct
  2. 2025Spatial Reasoning with Vision-Language Models in Ego-Centric Multi-View Scenes8 Sep
  3. 2025Reinforcing Spatial Reasoning in Vision-Language Models with Interwoven Thinking and Visual Drawing11 Jun
  4. 20253DRS: MLLMs Need 3D-Aware Representation Supervision for Scene Understanding2 Jun
  5. 2025MMSI-Bench: A Benchmark for Multi-Image Spatial Intelligence29 May
  6. 2025ViewSpatial-Bench: Evaluating Multi-perspective Spatial Localization in Vision-Language Models27 May
  7. 2025SSR: Enhancing Depth Perception in Vision-Language Models via Rationale-Guided Spatial Reasoning18 May
  8. 2025Perspective-Aware Reasoning in Vision-Language Models via Mental Imagery Simulation24 Apr
  9. 2025How to Enable LLM with 3D Capacity? A Survey of Spatial Reasoning in LLM8 Apr
  10. 2025SpaceR: Reinforcing MLLMs in Video Spatial Reasoning2 Apr
  11. 2025Open3D-VQA: A Benchmark for Comprehensive Spatial Reasoning with Multimodal Large Language Model in Open Space14 Mar
  12. 2025PointVLA: Injecting the 3D World into Vision-Language-Action Models10 Mar
3 more
  1. 2025SpatialCoT: Advancing Spatial Reasoning through Coordinate Alignment and Chain-of-Thought for Embodied Task Planning17 Jan
  2. 20243DSRBench: A Comprehensive 3D Spatial Reasoning Benchmark10 Dec
  3. 2024Reasoning Paths with Reference Objects Elicit Quantitative Spatial Reasoning in Large Vision-Language Models15 Sep

Other references · 51

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  2. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  3. 2025Qwen2.5-VL Technical Reportcited by 5,418
  4. 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,769
  5. 2025Group Sequence Policy Optimizationcited by 788
  6. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 763
  7. 2025Video-R1: Reinforcing Video Reasoning in MLLMscited by 464
  8. 2025DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learningcited by 398
  9. 2025LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Trainingcited by 233
  10. 2024From Explicit CoT to Implicit CoT: Learning to Internalize CoT Step by Stepcited by 225
  11. 2025Spatial-MLLM: Boosting MLLM Capabilities in Visual-based Spatial Intelligencecited by 206
  12. 2024Do Large Language Models Latently Perform Multi-Hop Reasoning?cited by 165
38 more
  1. 2025FutureSightDrive: Thinking Visually with Spatio-Temporal CoT for Autonomous Drivingcited by 165
  2. 2025OpenThinkIMG: Learning to Think with Images via Visual Tool Reinforcement Learningcited by 163
  3. 2025VLM-3R: Vision-Language Models Augmented with Instruction-Aligned 3D Reconstructioncited by 154
  4. 2024SpatialBot: Precise Spatial Understanding with Vision Language Modelscited by 146
  5. 2025LlamaV-o1: Rethinking Step-by-step Visual Reasoning in LLMscited by 145
  6. 2025From Flatland to Space: Teaching Vision-Language Models to Perceive and Reason in 3Dcited by 127
  7. 2025GRIT: Teaching MLLMs to Think with Imagescited by 121
  8. 2025Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokenscited by 120
  9. 2024SAT: Dynamic Spatial Aptitude Training for Multimodal Language Modelscited by 111
  10. 2025GPT4Scene: Understand 3D Scenes from Videos with Vision-Language Modelscited by 106
  11. 2025Latent Visual Reasoningcited by 99
  12. 2025MindCube: Spatial Mental Modeling from Limited Viewscited by 91
  13. 2025VTool-R1: VLMs Learn to Think with Images via Reinforcement Learning on Multimodal Tool Usecited by 84
  14. 2025UniVG-R1: Reasoning Guided Universal Visual Grounding with Reinforcement Learningcited by 83
  15. 2025JanusVLN: Decoupling Semantics and Spatiality with Dual Implicit Memory for Vision-Language Navigationcited by 80
  16. 2024Hopping Too Late: Exploring the Limitations of Large Language Models on Multi-Hop Queriescited by 78
  17. 2025MM-Spatial: Exploring 3D Spatial Understanding in Multimodal LLMscited by 72
  18. 2024OpenRLHF: An Easy-to-use, Scalable and High-performance RLHF Frameworkcited by 70
  19. 2024Toward Visual Grounding: A Surveycited by 52
  20. 2025VisRL: Intention-Driven Visual Perception via Reinforced Reasoningcited by 39
  21. 2024Align Anything: Training All-Modality Models to Follow Instructions with Language Feedbackcited by 38
  22. 2024SPHERE: Unveiling Spatial Blind Spots in Vision-Language Models Through Hierarchical Evaluationcited by 31
  23. 2025WonderTurbo: Generating Interactive 3D World in 0.72 Secondscited by 31
  24. 2023DepthSSC: Monocular 3D Semantic Scene Completion via Depth-Spatial Alignment and Voxel Adaptationcited by 30
  25. 2025Plan and Budget: Effective and Efficient Test-Time Scaling on Reasoning Large Language Modelscited by 30
  26. 2025EmbodieDreamer: Advancing Real2Sim2Real Transfer for Policy Training via Embodied World Modelingcited by 29
  27. 2025Holistic Evaluation of Multimodal LLMs on Spatial Intelligencecited by 28
  28. 2025ReconDreamer-RL: Enhancing Reinforcement Learning via Diffusion-based Scene Reconstructioncited by 27
  29. 2024SegLLM: Multi-round Reasoning Segmentationcited by 25
  30. 2025SIFThinker: Spatially-Aware Image Focus for Visual Reasoningcited by 25
  31. 2025Defining and Evaluating Visual Language Models' Basic Spatial Abilities: A Perspective from Psychometricscited by 21
  32. 2025Bridging the Editing Gap in LLMs: FineEdit for Precise and Targeted Text Modificationscited by 20
  33. 2025PromptSculptor: Multi-Agent Based Text-to-Image Prompt Optimizationcited by 19
  34. 2025Reasoning in Computer Vision: Taxonomy, Models, Tasks, and Methodologiescited by 17
  35. 2024YOLO-PPA based Efficient Traffic Sign Detection for Cruise Control in Autonomous Drivingcited by 16
  36. 2025Reasoning-Enhanced Domain-Adaptive Pretraining of Multimodal Large Language Models for Short Video Content Governancecited by 13
  37. 2025Audio-Enhanced Vision-Language Modeling with Latent Space Broadening for High Quality Data Expansioncited by 11
  38. 2025Neural Pruning for 3D Scene Reconstruction: Efficient NeRF Accelerationcited by 9

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 7; 2026: 51.

58 papers.

Sort:

By area:

  1. 2026GigaWorld-Policy: An Efficient Action-Centered World--Action Modelcited by 129
  2. 2025Scaling Spatial Intelligence with Multimodal Foundation Modelscited by 75
  3. 2025SwiftVLA: Unlocking Spatiotemporal Dynamics for Lightweight VLA Models at Minimal Overheadcited by 23
  4. 2026Think3D: Thinking with Space for Spatial Reasoningcited by 22
  5. 2026SCRIBE: Structured Mid-Level Supervision for Tool-Using Language Modelscited by 20
  6. 2025Mull-Tokens: Modality-Agnostic Latent Thinkingcited by 15
  7. 2026Thinking with Geometry: Active Geometry Integration for Spatial Reasoningcited by 14
  8. 2025Towards Spatial Trace with Reasoning in Vision-Language Models for Roboticscited by 12
  9. 2025SpatialDreamer: Incentivizing Spatial Reasoning via Active Mental Imagerycited by 10
  10. 2026Dual Latent Memory for Visual Multi-agent Systemcited by 9
  11. 2026OmniVideo-R1: Reinforcing Audio-visual Reasoning with Query Intention and Modality Attentioncited by 9
  12. 2026Spatial-TTT: Streaming Visual-based Spatial Intelligence with Test-Time Trainingcited by 8
38 more
  1. 2025Reward Auditor: Inference on Reward Modeling Suitability in Real-World Perturbed Scenarioscited by 7
  2. 2026RoboRouter: Training-Free Policy Routing for Robotic Manipulationcited by 6
  3. 2026Generation Models Know Space: Unleashing Implicit 3D Priors for Scene Understandingcited by 6
  4. 2026S2GR: Stepwise Semantic-Guided Reasoning in Latent Space for Generative Recommendationcited by 5
  5. 2026OpenSpatial: A Principled Data Engine for Empowering Spatial Intelligencecited by 4
  6. 2026Visual Latents Know More Than They Say: Unsilencing Latent Reasoning in MLLMscited by 4
  7. 2025Seeing through Imagination: Learning Scene Geometry via Implicit Spatial World Modelingcited by 3
  8. 2026EgoMind: Activating Spatial Cognition through Linguistic Reasoning in MLLMscited by 3
  9. 2026OpenWorldLib: A Unified Codebase and Definition of Advanced World Modelscited by 3
  10. 20264DThinker: Thinking with 4D Imagery for Dynamic Spatial Understandingcited by 3
  11. 2026Active Exploring like a Pigeon: Reinforcing Spatial Reasoning via Agentic Vision-Language Modelscited by 3
  12. 2026From Correspondence to Actions: Human-Like Multi-Image Spatial Reasoning in Multi-modal Large Language Modelscited by 2
  13. 2026How Far Are Vision-Language Models from Constructing the Real World? A Benchmark for Physical Generative Reasoningcited by 2
  14. 2026ProSR: Process-Shaped Spatial Reasoning for Reliable Chain-of-Thought in VLMscited by 2
  15. 2026Self-Evolving Neuro-Symbolic Skills for Tool-Augmented Spatial Reasoningcited by 2
  16. 2026Modeling Cross-vision Synergy for Unified Large Vision Modelcited by 1
  17. 20263D-RFT: Reinforcement Fine-Tuning for Video-based 3D Scene Understandingcited by 1
  18. 2026Enhancing MLLM Spatial Understanding via Active 3D Scene Exploration for Multi-Perspective Reasoningcited by 1
  19. 2026Thinking with Novel Views: A Systematic Analysis of Generative-Augmented Spatial Intelligencecited by 1
  20. 2026Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoningcited by 1
  21. 2026Ouroboros-Spatial: Closing the Data-Model Loop for Spatial Reasoningcited by 1
  22. 2026OmniView-Space: Reinforcing Spatial Reasoning via Multi-Perspective Spatial Mappingcited by 1
  23. 2026GeoAnchor: Collaborative Reasoning via Latent Decomposition for 3D Spatial Understandingcited by 1
  24. 2026Decoupling Stability and Plasticity for Multi-Modal Test-Time Adaptationcited by 0
  25. 2026GeoSense: Internalizing Geometric Necessity Perception for Multimodal Reasoningcited by 0
  26. 2026Token Warping Helps MLLMs Look from Nearby Viewpointscited by 0
  27. 2026MAG-3D: Multi-Agent Grounded Reasoning for 3D Understandingcited by 0
  28. 2026Dual-Pathway Geometry-Aware MLLM for Spatial Intelligencecited by 0
  29. 2026Reasmory: 3D Reconstruction as Explicit Memory for VLMs Spatial Reasoningcited by 0
  30. 2026SpatialSV: Internalizing Interpretable 3D Spatial Awareness in MLLMs via Task-Oriented Visual Supervisioncited by 0
  31. 2026OmniSpace: Efficient Geometry Awareness for Autonomous Vehicles MLLMscited by 0
  32. 2026VideoLatent: Video-Language Learning via Latent Self-Forcingcited by 0
  33. 2026OrientSAM: Mitigating Camera-Centric Shortcut in Multimodal Spatial Reasoning via Orientation-Aware Spatial Alignmentcited by 0
  34. 2026Advantage-Guided Gate: Reshaping Open-Ended Reasoning for Vision-Based Spatial Intelligencecited by 0
  35. 2026Chain of Spatial Thoughts: Modality-Agnostic Spatial Grounding for Vision Language Modelscited by 0
  36. 2026Multi-View Relational Distillation for Spatial Reasoning with Vision-Language Modelscited by 0
  37. 2026Unfold The World: Factorize 4D Properties in Reinforcing Spatial Reasoningcited by 0
  38. 2026Human-Centric Image Captioning with Subject-Centered Spatial Understandingcited by 0

50 most cited of 58 All 58 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 20263ViewSense: Spatial and Mental Perspective Reasoning from Orthographic Views in Vision-Language Models
  2. 2026Disentangling 3D Modeling from Spatial Reasoning
  3. 2025Reasoning Path and Latent State Analysis for Multi-view Visual Spatial Reasoning: A Cognitive Science Perspective
  4. 2026Make Geometry Matter for Spatial Reasoning
  5. 20263DThinkVLA: Endowing Vision-Language-Action Models with Latent 3D Priors via 3D-Thinking-Guided Co-training
  6. 2025Imagine in Space: Exploring the Frontier of Spatial Intelligence and Reasoning Efficiency in Vision Language Models
  7. 2025Spatial Understanding from Videos: Structured Prompts Meet Simulation Data
  8. 2026Boosting MLLM Spatial Reasoning with Geometrically Referenced 3D Scene Representations