mlsanityData to 7 Oct 2026

NarrativeFlow: Flow-Based Vision-Language-Action Model Using Robot Velocity Fields

arXiv 2610.00981 ↗First posted 1 Oct 2026cs.RO cs.CVcited by 0

On the map: Vision-language-action models → Human-video and latent-action pretraining also fits: Generalist VLA models and benchmarks

Abstract (arXiv)

We focus on language-conditioned flow-based manipulation, where robot flows (robot velocity fields) serve as embodiment-agnostic, motion-centric representations for leveraging data collected from multiple robot platforms. This task is crucial because language-conditioned manipulation is essential for practical robotic systems, yet scaling robot foundation models remains limited by the labor-intensive collection of embodiment-specific data. Existing methods either coarsely approximate robot flows with sparse keypoint displacements, or cannot handle language-conditioned manipulation. To address this limitation, we propose NarrativeFlow, which models robot flows as continuous velocity fields using a flow-matching formulation conditioned on language. Accordingly, NarrativeFlow generates robot flows that are physically consistent with real-world manipulation. To validate NarrativeFlow, we have conducted experiments on standard datasets for language-conditioned manipulation. The experimental results show that NarrativeFlow outperforms representative baseline methods on standard evaluation metrics. Furthermore, through real-world experiments, we show that NarrativeFlow achieves higher success rates than baseline methods across multiple manipulation tasks. The project page is available at https://shota0520.github.io/NarrativeFlow-project-page/

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Human-video and latent-action pretraining, one of 13 directions in Vision-language-action models. Builds on 58 of its 63 references that are in the map. First posted on arXiv 1 Oct 2026.

Builds on

58 of 63 references.

References · 58

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2022Scalable Diffusion Models with Transformerscited by 5,931
  3. 2022Flow Matching for Generative Modelingcited by 5,426
  4. 2025Qwen2.5-VL Technical Reportcited by 5,418
  5. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  6. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  7. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,007
  8. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,653
  9. 2019LXMERT: Learning Cross-Modality Encoder Representations from Transformerscited by 1,548
  10. 2025DINOv3cited by 1,297
  11. 2025Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Modelscited by 1,175
  12. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 801
38 more
  1. 2023BridgeData V2: A Dataset for Robot Learning at Scalecited by 634
  2. 2023CoTracker: It is Better to Track Togethercited by 516
  3. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 502
  4. 2025CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Modelscited by 440
  5. 2025AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systemscited by 390
  6. 2023Zero-Shot Robotic Manipulation with Pretrained Image-Editing Diffusion Modelscited by 264
  7. 2023Any-point Trajectory Modeling for Policy Learningcited by 207
  8. 2024RoboMIND: Benchmark on Multi-embodiment Intelligence Normative Data for Robot Manipulationcited by 166
  9. 2025RoboBrain: A Unified Brain Model for Robotic Manipulation from Abstract to Concretecited by 134
  10. 2024Flow as the Cross-Domain Manipulation Interfacecited by 126
  11. 2024CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videoscited by 122
  12. 2024Track2Act: Predicting Point Tracks from Internet Videos enables Generalizable Robot Manipulationcited by 116
  13. 2025A Survey on Vision-Language-Action Models: An Action Tokenization Perspectivecited by 96
  14. 2023RoboTAP: Tracking Arbitrary Points for Few-Shot Visual Imitationcited by 88
  15. 2025Vision-Language-Action Models for Robotics: A Review Towards Real-World Applicationscited by 82
  16. 2024General Flow as Foundation Affordance for Scalable Robot Learningcited by 80
  17. 2026PointWorld: Scaling 3D World Models for In-The-Wild Robotic Manipulationcited by 74
  18. 2025InstructVLA: Vision-Language-Action Instruction Tuning from Understanding to Manipulationcited by 71
  19. 2025MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Modelscited by 69
  20. 2025VidBot: Learning Generalizable 3D Actions from In-the-Wild 2D Human Videos for Zero-Shot Robotic Manipulationcited by 46
  21. 2024RT-Sketch: Goal-Conditioned Imitation Learning from Hand-Drawn Sketchescited by 41
  22. 2024FLIP: Flow-Centric Generative Planning as General-Purpose Manipulation World Modelcited by 41
  23. 2025RoboEngine: Plug-and-Play Robot Data Augmentation with Semantic Robot Segmentation and Background Generationcited by 40
  24. 2023PDPP: Projected Diffusion for Procedure Planning in Instructional Videoscited by 36
  25. 2023Learning Object State Changes in Videos: An Open-World Perspectivecited by 36
  26. 2024G3Flow: Generative 3D Semantic Flow for Pose-aware and Generalizable Object Manipulationcited by 35
  27. 2025Evo-1: Lightweight Vision-Language-Action Model with Preserved Semantic Alignmentcited by 35
  28. 2023FlowBot++: Learning Generalized Articulated Objects Manipulation via Articulation Projectioncited by 34
  29. 2023Goal Representations for Instruction Following: A Semi-Supervised Language Interface to Controlcited by 32
  30. 2025HiF-VLA: Hindsight, Insight and Foresight through Motion Representation for Vision-Language-Action Modelscited by 30
  31. 2024SCHEMA: State CHangEs MAtter for Procedure Planning in Instructional Videoscited by 24
  32. 2025ViPRA: Video Prediction for Robot Actionscited by 24
  33. 2024Tra-MoE: Learning Trajectory Prediction Model from Multiple Domains for Adaptive Policy Conditioningcited by 18
  34. 2025Gripper Keypose and Object Pointflow as Interfaces for Bimanual Robotic Manipulationcited by 16
  35. 2023Skip-Plan: Procedure Planning in Instructional Videos via Condensed Action Space Learningcited by 14
  36. 2024All Robots in One: A New Standard and Unified Dataset for Versatile, General-Purpose Embodied Agentscited by 14
  37. 2023Learning Language-Conditioned Deformable Object Manipulation with Graph Dynamicscited by 11
  38. 2025Generating 6DoF Object Manipulation Trajectories from Action Description in Egocentric Visioncited by 11

50 most cited of 58 All 58 (JSON)

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 20253DFlowAction: Learning Cross-Embodiment Manipulation from 3D Flow World Model
  2. 2026VLAFlow: A Unified Training Framework for Vision-Language-Action Models via Co-training and Future Latent Alignment
  3. 2026Task-Prototype Guided Flow Matching for Few-Shot Generalization in Vision-Language Robot Manipulation
  4. 2025ManiFlow: A General Robot Manipulation Policy via Consistency Flow Training
  5. 2023Bridging Language and Action: A Survey of Language-Conditioned Robot Manipulation
  6. 2026RoboInter: A Holistic Intermediate Representation Suite Towards Robotic Manipulation
  7. 2024Affordance-based Robot Manipulation with Flow Matching
  8. 2026ProbeFlow: Training-Free Adaptive Flow Matching for Vision-Language-Action Models