mlsanityData to 14 Oct 2026

GeoAlign: Beyond Semantics with State-Guided Spatial Alignment in VLA Models

arXiv 2606.03240 ↗First posted 2 Jun 2026cs.ROcited by 1

On the map: Vision-language-action models → 3D and spatial-aware VLA

Abstract (arXiv)

Current Vision-Language-Action (VLA) models often optimize for semantic grounding, whereas executable manipulation requires geometry-aware spatial alignment. We introduce GeoAlign, a state-guided spatial alignment architecture for VLA policy learning. Offline robot-domain RGB-D supervision post-trains an RGB geometry branch to produce Geometry-Enhanced Post-Trained (GEP) features; the depth head is then discarded, so policy training and rollout use RGB, language, and proprioceptive state. State-generated queries attend to the GEP grid to produce eight compact geometry tokens for action prediction. GeoAlign achieves 99.0% on LIBERO, 85.3% across three SimplerEnv-Fractal task families, and 78.8% on eight real-world ALOHA tasks, with ablations supporting the combined recipe of geometry post-training and state-guided querying on Isaac-GR00T N1.6-3B. Project website: https://chenyizhi123.github.io/geoalign-project/.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in 3D and spatial-aware VLA, one of 13 directions in Vision-language-action models. Builds on 48 of its 49 references that are in the map. First posted on arXiv 2 Jun 2026.

Builds on

48 of 49 references.

Closest ancestors · 7

Same direction, newest first

  1. 2026SG-VLA: Learning Spatially-Grounded Vision-Language-Action Models for Mobile Manipulation24 Mar
  2. 2025QDepth-VLA: Quantized Depth Prediction as Auxiliary Supervision for Vision-Language-Action Models16 Oct
  3. 2025DepthVLA: Enhancing Vision-Language-Action Models with Depth-Aware Spatial Reasoning15 Oct
  4. 2025GeoVLA: Empowering 3D Representations in Vision-Language-Action Models12 Aug
  5. 2025BridgeVLA: Input-Output Alignment for Efficient 3D Manipulation Learning with Vision-Language Models9 Jun
  6. 20253D CAVLA: Leveraging Depth and 3D Context to Generalize Vision Language Action Models for Unseen Tasks9 May
  7. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Model27 Jan

Other references · 41

  1. 2023BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Modelscited by 6,137
  2. 2022Scalable Diffusion Models with Transformerscited by 6,054
  3. 2022Flow Matching for Generative Modelingcited by 5,702
  4. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 3,080
  5. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  6. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,838
  7. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,407
  8. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,061
  9. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,761
  10. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,568
  11. 2024Octo: An Open-Source Generalist Robot Policycited by 1,484
  12. 2024Depth Anything V2cited by 1,422
29 more
  1. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,410
  2. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,225
  3. 2021Perceiver: General Perception with Iterative Attentioncited by 1,104
  4. 2020Object-Centric Learning with Slot Attentioncited by 876
  5. 2023VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Modelscited by 750
  6. 2024SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilitiescited by 706
  7. 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulationcited by 691
  8. 2021CLIPort: What and Where Pathways for Robotic Manipulationcited by 622
  9. 2022Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulationcited by 551
  10. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 495
  11. 2025CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Modelscited by 451
  12. 2024CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulationcited by 422
  13. 2025WorldVLA: Towards Autoregressive Action World Modelcited by 367
  14. 2020Transporter Networks: Rearranging the Visual World for Robotic Manipulationcited by 344
  15. 2024RoboPoint: A Vision-Language Model for Spatial Affordance Prediction for Roboticscited by 251
  16. 2023RVT: Robotic View Transformer for 3D Object Manipulationcited by 225
  17. 2025DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledgecited by 213
  18. 2025MolmoAct: Action Reasoning Models that can Reason in Spacecited by 197
  19. 2025Magma: A Foundation Model for Multimodal AI Agentscited by 139
  20. 2023Act3D: 3D Feature Field Transformers for Multi-Task Robotic Manipulationcited by 138
  21. 2024RoboSpatial: Teaching Spatial Understanding to 2D and 3D Vision-Language Models for Roboticscited by 136
  22. 2019ClearGrasp: 3D Shape Estimation of Transparent Objects for Manipulationcited by 124
  23. 2026RISE: Self-Improving Robot Policy with Compositional World Modelcited by 40
  24. 2026Masked Depth Modeling for Spatial Perceptioncited by 28
  25. 2025Goal-VLA: Image-Generative VLMs as Object-Centric World Models Empowering Zero-shot Robot Manipulationcited by 15
  26. 2025AnyPlace: Learning Generalized Object Placement for Robot Manipulationcited by 11
  27. 2026STARRY: Spatial-Temporal Action-Centric World Modeling for Robotic Manipulationcited by 10
  28. 2023Bridging the Sim-to-Real Gap with Dynamic Compliance Tuning for Industrial Insertioncited by 9
  29. 2024Stable Object Placement Planning From Contact Point Robustnesscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026An Action Is Worth One Patch: Unified World-Action Modeling with PatchWAMcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Mind-VLA: Instruction-Aware Spatial Representation Alignment for Vision-Language-Action Models
  2. 2025Spatial-Aware VLA Pretraining through Visual-Physical Alignment from Human Videos
  3. 2025Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Model
  4. 2026AWM-VLA: AlignedWorld Modeling for Efficient and Explainable Vision-Language-Action Policies
  5. 2026GeoBridge-VLA: Geometry-Aware Residual Adaptation for Vision-Language-Action Models
  6. 2025GeoPredict: Leveraging Predictive Kinematics and 3D Gaussian Geometry for Precise VLA Manipulation
  7. 2025GeoAware-VLA: Implicit Geometry Aware Vision-Language-Action Model
  8. 2026VEGA: Visual Encoder Grounding Alignment for Spatially-Aware Vision-Language-Action Models