mlsanityData to 7 Oct 2026

F1: A Vision-Language-Action Model Bridging Understanding and Generation to Actions

arXiv 2509.06951 ↗First posted 8 Sep 2025cs.RO cs.CVcited by 85

On the map: Vision-language-action models → World model-based VLA

Abstract (arXiv)

Executing language-conditioned tasks in dynamic visual environments remains a central challenge in embodied AI. Existing Vision-Language-Action (VLA) models predominantly adopt reactive state-to-action mappings, often leading to short-sighted behaviors and poor robustness in dynamic scenes. In this paper, we introduce F1, a pretrained VLA framework which integrates the visual foresight generation into decision-making pipeline. F1 adopts a Mixture-of-Transformer architecture with dedicated modules for perception, foresight generation, and control, thereby bridging understanding, generation, and actions. At its core, F1 employs a next-scale prediction mechanism to synthesize goal-conditioned visual foresight as explicit planning targets. By forecasting plausible future visual states, F1 reformulates action generation as a foresight-guided inverse dynamics problem, enabling actions that implicitly achieve visual goals. To endow F1 with robust and generalizable capabilities, we propose a three-stage training recipe on an extensive dataset comprising over 330k trajectories across 136 diverse tasks. This training scheme enhances modular reasoning and equips the model with transferable visual foresight, which is critical for complex and dynamic environments. Extensive evaluations on real-world tasks and simulation benchmarks demonstrate F1 consistently outperforms existing approaches, achieving substantial gains in both task success rate and generalization ability.

Key facts. Cited by 85 papers in the map (data through 7 Oct 2026). Placed in World model-based VLA, one of 13 directions in Vision-language-action models. Builds on 41 of its 53 references that are in the map. First posted on arXiv 8 Sep 2025.

Builds on

41 of 53 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge6 Jul

Other references · 40

  1. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,581
  2. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,048
  3. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  4. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  5. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,007
  6. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  7. 2019Root Mean Square Layer Normalizationcited by 1,425
  8. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  9. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,167
  10. 2024Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Predictioncited by 1,005
  11. 2025Emerging Properties in Unified Multimodal Pretrainingcited by 855
  12. 2024Emu3: Next-Token Prediction is All You Needcited by 761
28 more
  1. 2022Autoregressive Image Generation using Residual Quantizationcited by 706
  2. 2025FAST: Efficient Action Tokenization for Vision-Language-Action Modelscited by 643
  3. 2023Learning Universal Policies via Text-Guided Video Generationcited by 506
  4. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 502
  5. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 484
  6. 2024Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Modelcited by 468
  7. 2025CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Modelscited by 440
  8. 2025UniVLA: Learning to Act Anywhere with Task-centric Latent Actionscited by 386
  9. 2025BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Datasetcited by 363
  10. 2023Chameleon: Plug-and-Play Compositional Reasoning with Large Language Modelscited by 352
  11. 2024Video Prediction Policy: A Generalist Robot Policy with Predictive Visual Representationscited by 279
  12. 2023Zero-Shot Robotic Manipulation with Pretrained Image-Editing Diffusion Modelscited by 264
  13. 2025Unified World Models: Coupling Video and Action Diffusion for Pretraining on Large Robotic Datasetscited by 218
  14. 2025VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learningcited by 166
  15. 2024Predictive Inverse Dynamics Models are Scalable Learners for Robotic Manipulationcited by 159
  16. 2024Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Modelscited by 151
  17. 2025Genie Envisioner: A Unified World Foundation Platform for Robotic Manipulationcited by 129
  18. 2025Unified Vision-Language-Action Modelcited by 128
  19. 2025GraspVLA: a Grasping Foundation Model Pre-trained on Billion-scale Synthetic Action Datacited by 120
  20. 2025Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generationcited by 108
  21. 2024GRAPE: Generalizing Robot Policy via Preference Alignmentcited by 106
  22. 2024What Matters in Building Vision-Language-Action Models for Generalist Robotscited by 88
  23. 2025InstructVLA: Vision-Language-Action Instruction Tuning from Understanding to Manipulationcited by 71
  24. 2025Hume: Introducing System-2 Thinking in Visual-Language-Action Modelcited by 56
  25. 2025FlowVLA: Visual Chain of Thought-based Motion Reasoning for Vision-Language-Action Modelscited by 50
  26. 2024FLIP: Flow-Centric Generative Planning as General-Purpose Manipulation World Modelcited by 41
  27. 2025EO-1: An Open Unified Embodied Foundation Model for General Robot Controlcited by 30
  28. 2025CronusVLA: Towards Efficient and Robust Manipulation via Multi-Frame Vision-Language-Action Modelingcited by 24

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 12; 2026: 73.

85 papers.

Sort:

By area:

  1. 2025Motus: A Unified Latent Action World Modelcited by 231
  2. 2026Cosmos 3: Omnimodal World Models for Physical AIcited by 137
  3. 2026VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Modelcited by 104
  4. 2026ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learningcited by 94
  5. 2025InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policycited by 83
  6. 2026Being-H0.5: Scaling Human-Centric Robot Learning for Cross-Embodiment Generalizationcited by 61
  7. 2026InternVLA-A1: Unifying Understanding, Generation and Action for Robotic Manipulationcited by 48
  8. 2026World Action Models: The Next Frontier in Embodied AIcited by 46
  9. 2026LaWAM: Latent World Action Models for Efficient Dynamics-Aware Robot Policiescited by 46
  10. 2026World Model for Robot Learning: A Comprehensive Surveycited by 42
  11. 2025Unified Diffusion VLA: Vision-Language-Action Model via Joint Discrete Denoising Diffusion Processcited by 37
  12. 2026ImageWAM: Do World Action Models Really Need Video Generation, or Just Image Editing?cited by 33
38 more
  1. 2026MIND: Benchmarking Memory Consistency and Action Control in World Modelscited by 31
  2. 2026ACoT-VLA: Action Chain-of-Thought for Vision-Language-Action Modelscited by 30
  3. 2026Native Video-Action Pretraining for Generalizable Robot Controlcited by 30
  4. 2025VLA-Arena: An Open-Source Framework for Benchmarking Vision-Language-Action Modelscited by 28
  5. 2026ABot-M0.5: Unified Mobility-and-Manipulation World Action Modelcited by 28
  6. 2026BagelVLA: Enhancing Long-Horizon Manipulation via Interleaved Vision-Language-Action Generationcited by 25
  7. 2026OA-WAM: Object-Addressable World Action Model for Robust Robot Manipulationcited by 25
  8. 2026LangForce: Bayesian Decomposition of Vision Language Action Models via Latent Action Queriescited by 24
  9. 2026DualCoT-VLA: Visual-Linguistic Chain of Thought via Parallel Reasoning for Vision-Language-Action Modelscited by 20
  10. 2026Latent Reasoning VLA: Latent Thinking and Prediction for Vision-Language-Action Modelscited by 19
  11. 2026When to Trust Imagination: Adaptive Action Execution for World Action Modelscited by 18
  12. 2025AsyncVLA: Asynchronous Flow Matching for Vision-Language-Action Modelscited by 16
  13. 2026FutureVLA: Joint Visuomotor Prediction for Vision-Language-Action Modelcited by 14
  14. 2025ManualVLA: A Unified VLA Model for Chain-of-Thought Manual Generation and Robotic Manipulationcited by 13
  15. 2026HALO: A Unified Vision-Language-Action Model for Embodied Multimodal Chain-of-Thought Reasoningcited by 13
  16. 2026World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetrycited by 13
  17. 2026World-Value-Action Model: Implicit Planning for Vision-Language-Action Systemscited by 13
  18. 2025Mantis: A Versatile Vision-Language-Action Model with Disentangled Visual Foresightcited by 12
  19. 2026RoboInter: A Holistic Intermediate Representation Suite Towards Robotic Manipulationcited by 12
  20. 2026VLANeXt: Recipes for Building Strong VLA Modelscited by 12
  21. 2026DynVLA: Learning World Dynamics for Action Reasoning in Autonomous Drivingcited by 11
  22. 2026LaMP: Learning Vision-Language-Action Policy with 3D Scene Flow as Latent Motion Priorcited by 11
  23. 2025UniUGP: Unifying Understanding, Generation, and Planing For End-to-end Autonomous Drivingcited by 10
  24. 2026MMaDA-VLA: Large Diffusion Vision-Language-Action Model with Unified Multi-Modal Instruction and Generationcited by 10
  25. 2026STARRY: Spatial-Temporal Action-Centric World Modeling for Robotic Manipulationcited by 10
  26. 2026Metis: A Generalizable and Efficient World-Action Model for Autonomous Driving and Urban Navigationcited by 10
  27. 2026T-Rex: Tactile-Reactive Dexterous Manipulationcited by 10
  28. 2026World Action Models: A Surveycited by 10
  29. 2026DeltaVLA: Prior-Guided Vision-Language-Action Models via World Knowledge Variationcited by 9
  30. 2026Flex-π: A Multi-Stream World-Action Model with Compute Flexibilitycited by 9
  31. 2026ST4VLA: Spatially Guided Training for Vision-Language-Action Modelscited by 8
  32. 2026AffordanceVLA: A Vision-Language-Action Model Empowering Action Generation through Affordance-Aware Understandingcited by 8
  33. 2026HiMem-WAM: Hierarchical Memory-Gated World Action Models for Robotic Manipulationcited by 8
  34. 2025Unified Embodied VLM Reasoning with Robotic Action via Autoregressive Discretized Pre-trainingcited by 7
  35. 2026μVLA: On Recurrent Memory for Partially Observable Manipulation in VLA Modelscited by 7
  36. 2026OFlow: Injecting Object-Aware Temporal Flow Matching for Robust Robotic Manipulationcited by 6
  37. 2026Libra-VLA: Achieving Learning Equilibrium via Asynchronous Coarse-to-Fine Dual-Systemcited by 6
  38. 2026PRTS: A Primitive Reasoning and Tasking System via Contrastive Representationscited by 6

50 most cited of 85 All 85 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026ForeAct: Steering Your VLA with Efficient Visual Foresight Planning
  2. 2026FoMoVLA: Bridging Visual Foresight and Motion Guidance for Vision-Language-Action Models
  3. 2025Unifying Perception and Action: A Hybrid-Modality Pipeline with Implicit Visual Chain-of-Thought for Robotic Action Generation
  4. 2026PearlVLA: Progressive Embodied Action-Plan Refinement in Latent Space
  5. 2025HiF-VLA: Hindsight, Insight and Foresight through Motion Representation for Vision-Language-Action Models
  6. 2026Afford-VLA: Action-Aligned Visual Planning via Internalized Affordance
  7. 2026Anticipate Before Acting: Future-State-Conditioned Vision-Language Navigation
  8. 2026DIAL: Decoupling Intent and Action via Latent World Modeling for End-to-End VLA