mlsanityData to 7 Oct 2026

Dita: Scaling Diffusion Transformer for Generalist Vision-Language-Action Policy

arXiv 2503.19757 ↗First posted 25 Mar 2025cs.RO cs.CVcited by 73

On the map: Vision-language-action models → Efficient VLA inference and adaptation

also in 1 more
Abstract (arXiv)

While recent vision-language-action models trained on diverse robot datasets exhibit promising generalization capabilities with limited in-domain data, their reliance on compact action heads to predict discretized or continuous actions constrains adaptability to heterogeneous action spaces. We present Dita, a scalable framework that leverages Transformer architectures to directly denoise continuous action sequences through a unified multimodal diffusion process. Departing from prior methods that condition denoising on fused embeddings via shallow networks, Dita employs in-context conditioning -- enabling fine-grained alignment between denoised actions and raw visual tokens from historical observations. This design explicitly models action deltas and environmental nuances. By scaling the diffusion action denoiser alongside the Transformer's scalability, Dita effectively integrates cross-embodiment datasets across diverse camera perspectives, observation scenes, tasks, and action spaces. Such synergy enhances robustness against various variances and facilitates the successful execution of long-horizon tasks. Evaluations across extensive benchmarks demonstrate state-of-the-art or comparative performance in simulation. Notably, Dita achieves robust real-world adaptation to environmental variances and complex long-horizon tasks through 10-shot finetuning, using only third-person camera inputs. The architecture establishes a versatile, lightweight and open-source baseline for generalist robot policy learning. Project Page: https://robodita.github.io.

Key facts. Cited by 73 papers in the map (data through 7 Oct 2026). Placed in Efficient VLA inference and adaptation, one of 13 directions in Vision-language-action models. Builds on 72 of its 81 references that are in the map. First posted on arXiv 25 Mar 2025.

Builds on

72 of 81 references.

References · 72

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  4. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  5. 2020Denoising Diffusion Implicit Modelscited by 9,655
  6. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
  7. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  8. 2023BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Modelscited by 6,069
  9. 2022Scalable Diffusion Models with Transformerscited by 5,931
  10. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Controlcited by 2,993
  11. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  12. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
38 more
  1. 2022InstructPix2Pix: Learning to Follow Image Editing Instructionscited by 2,591
  2. 2023InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Taskscited by 2,534
  3. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  4. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,286
  5. 2023PaLM-E: An Embodied Multimodal Language Modelcited by 2,037
  6. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,007
  7. 2023Consistency Modelscited by 1,778
  8. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  9. 2024Octo: An Open-Source Generalist Robot Policycited by 1,427
  10. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  11. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 849
  12. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 801
  13. 2024PaliGemma: A versatile 3B VLM for transfercited by 679
  14. 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulationcited by 670
  15. 2024Autoregressive Image Generation without Vector Quantizationcited by 644
  16. 2024Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusioncited by 634
  17. 2021CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Taskscited by 564
  18. 2022Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulationcited by 545
  19. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 502
  20. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 484
  21. 2023DynamiCrafter: Animating Open-domain Images with Video Diffusion Priorscited by 463
  22. 2023An Embodied Generalist Agent in 3D Worldcited by 355
  23. 2023Unleashing Large-Scale Video Generative Pre-training for Visual Robot Manipulationcited by 353
  24. 2023Vision-Language Foundation Models as Effective Robot Imitatorscited by 344
  25. 2023ManiSkill2: A Unified Benchmark for Generalizable Manipulation Skillscited by 313
  26. 2019Learning Latent Plans from Playcited by 310
  27. 2024GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulationcited by 290
  28. 20243D Diffuser Actor: Policy Diffusion with 3D Scene Representationscited by 278
  29. 2023Zero-Shot Robotic Manipulation with Pretrained Image-Editing Diffusion Modelscited by 264
  30. 2023Imitating Human Behaviour with Diffusion Modelscited by 260
  31. 2023NoMaD: Goal Masked Diffusion Policies for Navigation and Explorationcited by 234
  32. 2023Goal-Conditioned Imitation Learning using Score-based Diffusion Policiescited by 229
  33. 2023ViNT: A Foundation Model for Visual Navigationcited by 210
  34. 2020Language Conditioned Imitation Learning over Unstructured Datacited by 208
  35. 2024RT-H: Action Hierarchies Using Languagecited by 204
  36. 2021ManiSkill: Generalizable Manipulation Skill Benchmark with Large-Scale Demonstrationscited by 193
  37. 2025DexVLA: Vision-Language Model with Plug-In Diffusion Expert for General Robot Controlcited by 191
  38. 2023RoboAgent: Generalization and Efficiency in Robot Manipulation via Semantic Augmentations and Action Chunkingcited by 167

50 most cited of 72 All 72 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 29; 2026: 44.

73 papers.

Sort:

By area:

  1. 2026Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planningcited by 274
  2. 2025DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledgecited by 203
  3. 2025Interactive Post-Training for Vision-Language-Action Modelscited by 129
  4. 2025Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Modelcited by 125
  5. 2025Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policiescited by 90
  6. 2026DiT4DiT: Jointly Modeling Video Dynamics and Actions for Generalizable Robot Controlcited by 71
  7. 2025CogVLA: Cognition-Aligned Vision-Language-Action Model via Instruction-Driven Routing & Sparsificationcited by 70
  8. 2025GeoVLA: Empowering 3D Representations in Vision-Language-Action Modelscited by 68
  9. 2025FlowVLA: Visual Chain of Thought-based Motion Reasoning for Vision-Language-Action Modelscited by 50
  10. 2025Pure Vision Language Action (VLA) Models: A Comprehensive Surveycited by 46
  11. 2025VLA-OS: Structuring and Dissecting Planning Representations and Paradigms in Vision-Language-Action Modelscited by 34
  12. 2025Visual Embodied Brain: Let Multimodal Large Language Models See, Think, and Control in Spacescited by 33
38 more
  1. 2025Exploring the Limits of Vision-Language-Action Manipulations in Cross-task Generalizationcited by 31
  2. 2025RoboCerebra: A Large-scale Benchmark for Long-horizon Robotic Manipulation Evaluationcited by 31
  3. 2025CronusVLA: Towards Efficient and Robust Manipulation via Multi-Frame Vision-Language-Action Modelingcited by 24
  4. 2026Adaptive Action Chunking at Inference-time for Vision-Language-Action Modelscited by 24
  5. 2025Vlaser: Vision-Language-Action Model with Synergistic Embodied Reasoningcited by 22
  6. 2026PALM: Progress-Aware Policy Learning via Affordance Reasoning for Long-Horizon Robotic Manipulationcited by 20
  7. 2025Align-Then-stEer: Adapting the Vision-Language Action Models through Unified Latent Guidancecited by 18
  8. 2025Steering Vision-Language-Action Models as Anti-Exploration: A Test-Time Scaling Approachcited by 18
  9. 2025Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbonecited by 18
  10. 2025Parallels Between VLA Model Post-Training and Human Motor Learning: Progress, Challenges, and Trendscited by 17
  11. 2025Seeing to Act, Prompting to Specify: A Bayesian Factorization of Vision Language Action Policycited by 17
  12. 2025RoVer: Robot Reward Model as Test-Time Verifier for Vision-Language-Action Modelcited by 16
  13. 2025RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action Modelscited by 16
  14. 2025PixelVLA: Advancing Pixel-level Understanding in Vision-Language-Action Modelcited by 16
  15. 2025Balancing Signal and Variance: Adaptive Offline RL Post-Training for VLA Flow Modelscited by 14
  16. 2025FreqPolicy: Efficient Flow-based Visuomotor Policy via Frequency Consistencycited by 13
  17. 2026Stable Language Guidance for Vision-Language-Action Modelscited by 13
  18. 2026Any3D-VLA: Enhancing VLA Robustness via Diverse Point Cloudscited by 13
  19. 2026ROCKET: Residual-Oriented Multi-Layer Alignment for Spatially-Aware Vision-Language-Action Modelscited by 11
  20. 2026DeltaVLA: Prior-Guided Vision-Language-Action Models via World Knowledge Variationcited by 9
  21. 2025Grounding Actions in Camera Space: Observation-Centric Vision-Language-Action Policycited by 8
  22. 2026Chain of World: World Model Thinking in Latent Motioncited by 8
  23. 2026ProgressVLA: Progress-Guided Diffusion Policy for Vision-Language Robotic Manipulationcited by 8
  24. 2026Point Tracking Improves World Action Modelscited by 7
  25. 2026Generation Models Know Space: Unleashing Implicit 3D Priors for Scene Understandingcited by 6
  26. 2026ConsisVLA-4D: Advancing Spatiotemporal Consistency in Efficient 3D-Perception and 4D-Reasoning for Robotic Manipulationcited by 6
  27. 2026Sparse ActionGen: Accelerating Diffusion Policy with Real-time Pruningcited by 5
  28. 2026ReViP: Mitigating False Completion in Vision-Language-Action Models with Vision-Proprioception Rebalancecited by 5
  29. 2026Cross-Hand Latent Representation for Vision-Language-Action Modelscited by 5
  30. 2026Retrieve-then-Steer: Online Success Memory for Test-Time Adaptation of Generative VLAscited by 5
  31. 2026UniSteer: Unified Noise Steering for Efficient Human-Guided VLA Adaptationcited by 5
  32. 2026Escaping the Diversity Trap in Robotic Manipulation via Anchor-Centric Adaptationcited by 4
  33. 2026FAR: Failure-Aware Retry for Test-Time Recovery and Continual Policy Improvementcited by 4
  34. 2025E0: Enhancing Generalization and Fine-Grained Control in VLA Models via Tweedie Discrete Diffusioncited by 3
  35. 2026Policy-Guided World Model Planning for Language-Conditioned Visual Navigationcited by 3
  36. 2026Learning Transferable Dynamics Priors from Action to World Modelingcited by 3
  37. 2025FORGE-Tree: Diffusion-Forcing Tree Search for Long-Horizon Robot Manipulationcited by 2
  38. 2025BLM1: A Boundless Large Model for Cross-Space, Cross-Task, and Cross-Embodiment Learningcited by 2

50 most cited of 73 All 73 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024Diffusion Transformer Policy
  2. 2026StairVLA: Stage-Aware Hierarchical Action Generation for Vision-Language-Action Models
  3. 2026Discrete Forcing: Infusing Discrete Guidance into Continuous Denoising for Few-Step Action Experts
  4. 2024Prediction with Action: Visual Policy Learning via Joint Denoising Process
  5. 2025Dual-Stream Diffusion for World-Model Augmented Vision-Language-Action Model
  6. 2026Beyond Future Prediction: Denoising as Generative Adaptation for Robot Control
  7. 2026dVLA-RL: Reinforcement Learning over Denoising Trajectories for Discrete Diffusion Vision-Language-Action Models
  8. 2026MMaDA-VLA: Large Diffusion Vision-Language-Action Model with Unified Multi-Modal Instruction and Generation