mlsanityData to 7 Oct 2026

Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success

arXiv 2502.19645 ↗First posted 27 Feb 2025cs.RO cs.AI cs.CVcited by 844

On the map: LLM reasoning → Spatial and embodied reasoning

also in 2 more
Abstract (arXiv)

Recent vision-language-action models (VLAs) build upon pretrained vision-language models and leverage diverse robot datasets to demonstrate strong task execution, language following ability, and semantic generalization. Despite these successes, VLAs struggle with novel robot setups and require fine-tuning to achieve good performance, yet how to most effectively fine-tune them is unclear given many possible strategies. In this work, we study key VLA adaptation design choices such as different action decoding schemes, action representations, and learning objectives for fine-tuning, using OpenVLA as our representative base model. Our empirical analysis informs an Optimized Fine-Tuning (OFT) recipe that integrates parallel decoding, action chunking, a continuous action representation, and a simple L1 regression-based learning objective to altogether improve inference efficiency, policy performance, and flexibility in the model's input-output specifications. We propose OpenVLA-OFT, an instantiation of this recipe, which sets a new state of the art on the LIBERO simulation benchmark, significantly boosting OpenVLA's average success rate across four task suites from 76.5% to 97.1% while increasing action generation throughput by 26×. In real-world evaluations, our fine-tuning recipe enables OpenVLA to successfully execute dexterous, high-frequency control tasks on a bimanual ALOHA robot and outperform other VLAs (π0 and RDT-1B) fine-tuned using their default recipes, as well as strong imitation learning policies trained from scratch (Diffusion Policy and ACT) by up to 15% (absolute) in average success rate. We release code for OFT and pretrained model checkpoints at https://openvla-oft.github.io/.

Key facts. Cited by 844 papers in the map (data through 7 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 50 of its 66 references that are in the map. First posted on arXiv 27 Feb 2025.

Builds on

50 of 66 references.

References · 50

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  4. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  5. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  6. 2020Denoising Diffusion Implicit Modelscited by 9,655
  7. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  8. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  9. 2022Flow Matching for Generative Modelingcited by 5,426
  10. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,243
  11. 2016Gaussian Error Linear Units (GELUs)cited by 3,431
  12. 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,048
38 more
  1. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  2. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  3. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  4. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,302
  5. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  6. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,286
  7. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  8. 2024Octo: An Open-Source Generalist Robot Policycited by 1,427
  9. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  10. 2022Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agentscited by 1,029
  11. 2022DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Modelscited by 924
  12. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 849
  13. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 801
  14. 2023VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Modelscited by 733
  15. 2022R3M: A Universal Visual Representation for Robot Manipulationcited by 678
  16. 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulationcited by 670
  17. 2025FAST: Efficient Action Tokenization for Vision-Language-Action Modelscited by 643
  18. 2023BridgeData V2: A Dataset for Robot Learning at Scalecited by 634
  19. 2022ProgPrompt: Generating Situated Robot Task Plans using Large Language Modelscited by 598
  20. 2022LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Modelscited by 459
  21. 2022VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Trainingcited by 384
  22. 2023An Embodied Generalist Agent in 3D Worldcited by 355
  23. 2023Vision-Language Foundation Models as Effective Robot Imitatorscited by 344
  24. 2024Prismatic VLMs: Investigating the Design Space of Visually-Conditioned Language Modelscited by 336
  25. 2024TinyVLA: Towards Fast, Data-Efficient Vision-Language-Action Models for Robotic Manipulationcited by 331
  26. 20243D-VLA: A 3D Vision-Language-Action Generative World Modelcited by 320
  27. 2023Where are we in the search for an Artificial Visual Cortex for Embodied Intelligence?cited by 215
  28. 2022CoWs on Pasture: Baselines and Benchmarks for Language-Driven Zero-Shot Object Navigationcited by 194
  29. 2023LIV: Language-Image Representations and Rewards for Robotic Controlcited by 180
  30. 2023Language-Driven Representation Learning for Roboticscited by 163
  31. 2024Predictive Inverse Dynamics Models are Scalable Learners for Robotic Manipulationcited by 159
  32. 2023Open-World Object Manipulation using Pre-trained Vision-Language Modelscited by 155
  33. 2024Multimodal Diffusion Transformer: Learning Versatile Behavior from Multimodal Goalscited by 131
  34. 2024ALOHA Unleashed: A Simple Recipe for Robot Dexteritycited by 128
  35. 2024Yell At Your Robot: Improving On-the-Fly from Language Correctionscited by 123
  36. 2024Manipulate-Anything: Automating Real-World Robots using Vision-Language Modelscited by 82
  37. 2024An Interactive Agent Foundation Modelcited by 25
  38. 2024Language Conditioned Multi-Finger Dexterous Manipulation Enabled by Physical Compliance and Switching of Controllerscited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 205; 2026: 639.

844 papers.

Sort:

By area:

  1. 2025RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulationcited by 477
  2. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  3. 2025WorldVLA: Towards Autoregressive Action World Modelcited by 351
  4. 2026Causal World Modeling for Robot Controlcited by 289
  5. 2025X-VLA: Soft-Prompted Transformer as Scalable Cross-Embodiment Vision-Language-Action Modelcited by 285
  6. 2025LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Modelscited by 278
  7. 2026Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planningcited by 274
  8. 2025Real-Time Execution of Action Chunking Flow Policiescited by 199
  9. 2025MolmoAct: Action Reasoning Models that can Reason in Spacecited by 189
  10. 2025MemoryVLA: Perceptual-Cognitive Memory in Vision-Language-Action Models for Robotic Manipulationcited by 183
  11. 2025A Survey of Reinforcement Learning for Large Reasoning Modelscited by 175
  12. 2025Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Bettercited by 142
38 more
  1. 2025SimpleVLA-RL: Scaling VLA Training via Reinforcement Learningcited by 141
  2. 2025VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Modelcited by 140
  3. 2025Interactive Post-Training for Vision-Language-Action Modelscited by 129
  4. 2025Unified Vision-Language-Action Modelcited by 128
  5. 2025Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Modelcited by 125
  6. 2025NORA: A Small Open-Sourced Generalist Vision Language Action Model for Embodied Taskscited by 119
  7. 2025GR-3 Technical Reportcited by 110
  8. 2025FLARE: Robot Learning with Implicit World Modelingcited by 105
  9. 2026VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Modelcited by 104
  10. 2026StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developingcited by 103
  11. 2025What Can RL Bring to VLA Generalization? An Empirical Studycited by 102
  12. 2026ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learningcited by 94
  13. 2025Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policiescited by 90
  14. 2025PD-VLA: Accelerating Vision-Language-Action Model Integrated with Action Chunking via Parallel Decodingcited by 88
  15. 2025DexGraspVLA: A Vision-Language-Action Framework Towards General Dexterous Graspingcited by 87
  16. 2025ForceVLA: Enhancing VLA Models with a Force-aware MoE for Contact-rich Manipulationcited by 87
  17. 2025InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policycited by 83
  18. 2025Vision-Language-Action Models for Robotics: A Review Towards Real-World Applicationscited by 82
  19. 2025SAFE: Multitask Failure Detection for Vision-Language-Action Modelscited by 74
  20. 2025Vision-Language-Action (VLA) Models: Concepts, Progress, Applications and Challengescited by 72
  21. 2025A Survey on Vision-Language-Action Models for Autonomous Drivingcited by 72
  22. 2025Self-Improving Vision-Language-Action Models with Data Generation via Residual RLcited by 72
  23. 2025InstructVLA: Vision-Language-Action Instruction Tuning from Understanding to Manipulationcited by 71
  24. 2026DiT4DiT: Jointly Modeling Video Dynamics and Actions for Generalizable Robot Controlcited by 71
  25. 2025CogVLA: Cognition-Aligned Vision-Language-Action Model via Instruction-Driven Routing & Sparsificationcited by 70
  26. 2025RynnVLA-002: A Unified Vision-Language-Action and World Modelcited by 70
  27. 2025GeoVLA: Empowering 3D Representations in Vision-Language-Action Modelscited by 68
  28. 2025WMPO: World Model-based Policy Optimization for Vision-Language-Action Modelscited by 68
  29. 2026Being-H0.7: A Latent World-Action Model from Egocentric Videoscited by 68
  30. 2025VQ-VLA: Improving Vision-Language-Action Models via Scaling Vector-Quantized Action Tokenizerscited by 67
  31. 2025Large VLM-based Vision-Language-Action Models for Robotic Manipulation: A Surveycited by 66
  32. 2025A Vision-Language-Action-Critic Model for Robotic Real-World Reinforcement Learningcited by 66
  33. 2025VLA-RFT: Vision-Language-Action Reinforcement Fine-tuning with Verified Rewards in World Simulatorscited by 65
  34. 2025WorldEval: World Model as Real-World Robot Policies Evaluatorcited by 64
  35. 2025Scalable Vision-Language-Action Model Pretraining for Robotic Manipulation with Real-Life Human Activity Videoscited by 63
  36. 2026Being-H0.5: Scaling Human-Centric Robot Learning for Cross-Embodiment Generalizationcited by 61
  37. 2025villa-X: Enhancing Latent Action Modeling in Vision-Language-Action Modelscited by 58
  38. 2025World-Env: Leveraging World Model as a Virtual Environment for VLA Post-Trainingcited by 57

50 most cited of 844 All 844 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026FineVLA: Fine-Grained Instruction Alignment for Steerable Vision-Language-Action Policies
  2. 2026Adaptive Capacity Allocation for Vision Language Action Fine-tuning
  3. 2024Benchmarking Vision, Language, & Action Models on Robotic Learning Tasks
  4. 2026Rethinking Visual-Language-Action Model Scaling: Alignment, Mixture, and Regularization
  5. 2025Dual-Actor Fine-Tuning of VLA Models: A Talk-and-Tweak Human-in-the-Loop Approach
  6. 2025Mechanistic Finetuning of Vision-Language-Action Models via Few-Shot Demonstrations
  7. 2026Learning What to Say to Your VLA: Mostly Harmless Vision Language Action Model Steering
  8. 2025From Intention to Execution: Probing the Generalization Boundaries of Vision-Language-Action Models