mlsanityData to 14 Oct 2026

OmniVLA-RL: A Vision-Language-Action Model with Spatial Understanding and Online RL

arXiv 2604.17706 ↗First posted 20 Apr 2026cs.ROcited by 0

On the map: Vision-language-action models → Reinforcement learning and post-training

also in 1 more
Abstract (arXiv)

Visual-Language-Action (VLA) models represent a paradigm shift in embodied AI, yet existing frameworks often struggle with imprecise spatial perception, suboptimal multimodal fusion, and instability in reinforcement learning. To bridge these gaps, we propose OmniVLA-RL, a novel architecture that leverages a Mix-of-Transformers (MoT) design to synergistically integrate reasoning, spatial, and action experts. Furthermore, we introduce Flow-GSPO, which reformulates flow matching as a Stochastic Differential Equation (SDE) process and integrates it with Group Segmented Policy Optimization (GSPO) to enhance action precision and training robustness. Extensive evaluations on the LIBERO and LIBERO-Plus benchmarks demonstrate that OmniVLA-RL achieves decent overall performance and surpasses mainstream existing methods, effectively overcoming the fundamental limitations of current VLA models.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Reinforcement learning and post-training, one of 13 directions in Vision-language-action models. Builds on 49 of its 54 references that are in the map. First posted on arXiv 20 Apr 2026.

Builds on

49 of 54 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025π*0.6: a VLA That Learns From Experience18 Nov
  2. 2025VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning24 May

Other references · 47

  1. 2017Attention Is All You Needcited by 60,361
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  4. 2017Proximal Policy Optimization Algorithmscited by 15,215
  5. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  6. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  7. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  8. 2022Flow Matching for Generative Modelingcited by 5,702
  9. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,101
  10. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,761
  11. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,109
  12. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 3,080
35 more
  1. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,887
  2. 2015Trust Region Policy Optimizationcited by 2,580
  3. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,407
  4. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,224
  5. 2023Consistency Modelscited by 1,822
  6. 2025VGGT: Visual Geometry Grounded Transformercited by 1,539
  7. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,410
  8. 2025Group Sequence Policy Optimizationcited by 835
  9. 2023VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Modelscited by 750
  10. 2024SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilitiescited by 706
  11. 2025SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Roboticscited by 530
  12. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 519
  13. 2025π3: Permutation-Equivariant Visual Geometry Learningcited by 319
  14. 2022Rectified Flow: A Marginal Preserving Approach to Optimal Transportcited by 283
  15. 2024Simplifying, Stabilizing and Scaling Continuous-Time Consistency Modelscited by 283
  16. 2024Diffusion Policy Policy Optimizationcited by 280
  17. 2025SmolVLM: Redefining small and efficient multimodal modelscited by 237
  18. 2024PaliGemma 2: A Family of Versatile VLMs for Transfercited by 207
  19. 2025LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorizationcited by 179
  20. 2024Consistency Policy: Accelerated Visuomotor Policies via Consistency Distillationcited by 166
  21. 2024Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Modelscited by 163
  22. 2022M2BEV: Multi-Camera Joint 3D Detection and Segmentation with Unified Birds-Eye View Representationcited by 156
  23. 2025Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Modelcited by 134
  24. 2025DriveVLA-W0: World Models Amplify Data Scaling Law in Autonomous Drivingcited by 134
  25. 2024FlowPolicy: Enabling Fast and Robust 3D Flow-based Policy via Consistency Flow Matching for Robot Manipulationcited by 104
  26. 2025F1: A Vision-Language-Action Model Bridging Understanding and Generation to Actionscited by 86
  27. 2025Evo-0: Vision-Language-Action Model with Implicit Spatial Understandingcited by 58
  28. 2025G2VLM: Geometry Grounded Vision Language Model with Unified 3D Reconstruction and Spatial Reasoningcited by 44
  29. 2023Fast-BEV: A Fast and Strong Bird's-Eye View Perception Baselinecited by 40
  30. 2025CEED-VLA: Consistency Vision-Language-Action Model with Early-Exit Decodingcited by 38
  31. 2025From Spatial to Actions: Grounding Vision-Language-Action Model in Spatial Foundation Priorscited by 28
  32. 2017Policy Gradient Methods for Reinforcement Learning with Function Approximation and Action-Dependent Baselinescited by 20
  33. 2024Score-based generative models are provably robust: an uncertainty quantification perspectivecited by 15
  34. 2025Analyzing the Mechanism of Attention Collapse in VGGT from a Dynamics Perspectivecited by 6
  35. 2025Progressive Supernet Training for Efficient Visual Autoregressive Modelingcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025VLA-R1: Enhancing Reasoning in Vision-Language-Action Models
  2. 2025RLinf-VLA: A Unified and Efficient Framework for Reinforcement Learning of Vision-Language-Action Models
  3. 2026SA-VLA: Spatially-Aware Flow-Matching for Vision-Language-Action Reinforcement Learning
  4. 2025Vision-Language-Action (VLA) Models: Concepts, Progress, Applications and Challenges
  5. 2026StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing
  6. 2026AdaVLA: Adaptive Step Flow Matching for Training-free Acceleration of Vision-Language-Action Models
  7. 2025π_textttRL: Online RL Fine-tuning for Flow-based Vision-Language-Action Models
  8. 2026Towards Practical World Model-based Reinforcement Learning for Vision-Language-Action Models