mlsanityData to 7 Oct 2026

Boosting Vision-Language-Action Finetuning with Feasible Action Neighborhood Prior

arXiv 2604.01570 ↗First posted 2 Apr 2026cs.ROcited by 0

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

Abstract (arXiv)

In real-world robotic manipulation, states typically admit a neighborhood of near-equivalent actions. That is, for each state, there exist a feasible action neighborhood (FAN) rather than a single correct action, within which motions yield indistinguishable progress. However, prevalent VLA training methodologies are directly inherited from linguistic settings and do not exploit the FAN property, thus leading to poor generalization and low sample efficiency. To address this limitation, we introduce a FAN-guided regularizer that shapes the model's output distribution to align with the geometry of FAN. Concretely, we introduce a Gaussian prior that promotes locally smooth and unimodal predictions around the preferred direction and magnitude. In extensive experiments across both reinforced finetuning (RFT) and supervised finetuning (SFT), our method achieves significant improvement in sample efficiency, and success rate in both in-distribution and out-of-distribution (OOD) scenarios. By aligning with the intrinsic action tolerance of physical manipulation, FAN-guided regularization provides a principled and practical method for sample-efficient, and generalizable VLA adaptation.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reinforcement learning and post-training, one of 13 directions in Vision-language-action models. Builds on 52 of its 58 references that are in the map. First posted on arXiv 2 Apr 2026.

Builds on

52 of 58 references.

Closest ancestors · 11

Same direction, newest first

  1. 2025SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning11 Sep
  2. 2025CO-RFT: Efficient Fine-Tuning of Vision-Language-Action Models through Chunked Offline Reinforcement Learning4 Aug
  3. 2025TGRPO :Fine-tuning Vision-Language-Action Model via Trajectory-wise Group Relative Policy Optimization10 Jun
  4. 2025What Can RL Bring to VLA Generalization? An Empirical Study26 May
  5. 2025VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning24 May
  6. 2025Interactive Post-Training for Vision-Language-Action Models22 May
  7. 2025ReinboT: Amplifying Robot Visual-Language Manipulation with Reinforcement Learning12 May
  8. 2025ConRFT: A Reinforced Fine-tuning Method for VLA Models via Consistency Policy8 Feb
  9. 2025Improving Vision-Language-Action Model with Online Reinforcement Learning28 Jan
  10. 2024GRAPE: Generalizing Robot Policy via Preference Alignment28 Nov
  11. 2024Octo: An Open-Source Generalist Robot Policy20 May

Other references · 41

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  4. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  5. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  6. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  7. 2023Let's Verify Step by Stepcited by 3,598
  8. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,048
  9. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  10. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  11. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,302
  12. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
29 more
  1. 2023PaLM-E: An Embodied Multimodal Language Modelcited by 2,037
  2. 2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 1,916
  3. 2025π0.5: a Vision-Language-Action Model with Open-World Generalizationcited by 1,653
  4. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  5. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  6. 2022Code as Policies: Language Model Programs for Embodied Controlcited by 1,172
  7. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,167
  8. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 849
  9. 2025Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Successcited by 844
  10. 2023VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Modelscited by 733
  11. 2018Attention, Learn to Solve Routing Problems!cited by 646
  12. 2025SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Roboticscited by 507
  13. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 502
  14. 2025UniVLA: Learning to Act Anywhere with Task-centric Latent Actionscited by 386
  15. 2025WorldVLA: Towards Autoregressive Action World Modelcited by 351
  16. 2025LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Modelscited by 278
  17. 2024ManiSkill3: GPU Parallelized Robotics Simulation and Rendering for Generalizable Embodied AIcited by 216
  18. 2024RT-H: Action Hierarchies Using Languagecited by 204
  19. 2025HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Modelcited by 176
  20. 2025LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorizationcited by 166
  21. 2023Look Before You Leap: Unveiling the Power of GPT-4V in Robotic Vision-Language Planningcited by 161
  22. 2025OneTwoVLA: A Unified Vision-Language-Action Model with Adaptive Reasoningcited by 105
  23. 2025EfficientVLA: Training-Free Acceleration and Compression for Vision-Language-Action Modelscited by 89
  24. 2024Policy Agnostic RL: Offline RL and Online RL Fine-Tuning of Any Class and Backbonecited by 71
  25. 2025VQ-VLA: Improving Vision-Language-Action Models via Scaling Vector-Quantized Action Tokenizerscited by 67
  26. 2025Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Surveycited by 66
  27. 2024FLaRe: Achieving Masterful and Adaptive Robot Policies with Large-Scale Reinforcement Learning Fine-Tuningcited by 54
  28. 2025BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulationcited by 42
  29. 2025Parallels Between VLA Model Post-Training and Human Motor Learning: Progress, Challenges, and Trendscited by 17

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026FAN: Foresight Action Normalization for Continual Adaptation of Vision-Language-Action Models
  2. 2024CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
  3. 2026PriorVLA: Prior-Preserving Adaptation for Vision-Language-Action Models
  4. 2025E0: Enhancing Generalization and Fine-Grained Control in VLA Models via Tweedie Discrete Diffusion
  5. 2026ActionUNet: Improving Robustness of VLA Models with Efficient Multi-scale Fine-tuning
  6. 2026Reshaping Action Error Distributions for Reliable Vision-Language-Action Models
  7. 2025Align-Then-stEer: Adapting the Vision-Language Action Models through Unified Latent Guidance
  8. 2026Generalizable VLA Finetuning via Representation Anchoring and Language-Action Alignment