mlsanityData to 7 Oct 2026

Domain Arithmetic: One-Shot VLA Adaptation under Environmental Shifts

arXiv 2607.00666 ↗First posted 1 Jul 2026cs.RO cs.CV cs.LGcited by 0

On the map: Vision-language-action models → Generalist VLA models and benchmarks

Abstract (arXiv)

Vision-Language-Action (VLA) models often fail to perform the same learned tasks under environmental shifts, such as changes in camera pose and shifts to a different but similar robot (e.g., from Panda to UR5e). Adapting these models to the shifted environment (i.e., target domain) often requires training on multiple demonstrations for each task, which are costly to collect. To reduce the burden of data curation and training, we propose an analogy-based method that adapts VLA models under environmental shifts through weight vector arithmetic with domain-specific information addition, named Domain ARiThmetic (DART). Unlike prior approaches, DART requires collecting only a single demonstration, enabling efficient adaptation. To accurately isolate domain-specific information for addition, DART performs subspace alignment between singular components in weight vectors to filter out noisy components. In both simulated and real-world experiments, DART outperforms existing VLA adaptation methods in one-shot scenarios across diverse visual and embodiment shifts. Code is available at https://github.com/snumprlab/dart.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Generalist VLA models and benchmarks, one of 13 directions in Vision-language-action models. Builds on 73 of its 80 references that are in the map. First posted on arXiv 1 Jul 2026.

Builds on

73 of 80 references.

Closest ancestors · 8

Same direction, newest first

  1. 2026Action Hallucination in Generative Vision-Language-Action Models6 Feb
  2. 2025MergeVLA: Cross-Skill Model Merging Toward a Generalist Vision-Language-Action Agent24 Nov
  3. 2025π0.5: a Vision-Language-Action Model with Open-World Generalization22 Apr
  4. 2025A Taxonomy for Evaluating Generalist Robot Manipulation Policies3 Mar
  5. 2025Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success27 Feb
  6. 2024π0: A Vision-Language-Action Flow Model for General Robot Control31 Oct
  7. 2024OpenVLA: An Open-Source Vision-Language-Action Model13 Jun
  8. 2023RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control28 Jul

Other references · 65

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  2. 2023Visual Instruction Tuningcited by 7,792
  3. 2022Flow Matching for Generative Modelingcited by 5,426
  4. 2022Locating and Editing Factual Associations in GPTcited by 2,609
  5. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,338
  6. 2025GR00T N1: An Open Foundation Model for Generalist Humanoid Robotscited by 1,167
  7. 2022Editing Models with Task Arithmeticcited by 1,076
  8. 2022Mass-Editing Memory in a Transformercited by 976
  9. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 849
  10. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 801
  11. 2023TIES-Merging: Resolving Interference When Merging Modelscited by 795
  12. 2023Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunchcited by 679
38 more
  1. 2024PaliGemma: A versatile 3B VLM for transfercited by 679
  2. 2022R3M: A Universal Visual Representation for Robot Manipulationcited by 678
  3. 2025FAST: Efficient Action Tokenization for Vision-Language-Action Modelscited by 643
  4. 2023BridgeData V2: A Dataset for Robot Learning at Scalecited by 634
  5. 2024The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearningcited by 488
  6. 2024Evaluating Real-World Robot Manipulation Policies in Simulationcited by 484
  7. 2024GaLore: Memory-Efficient LLM Training by Gradient Low-Rank Projectioncited by 447
  8. 2023MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrationscited by 308
  9. 2022Behavior Transformers: Cloning k modes with one stonecited by 299
  10. 2025π*0.6: a VLA That Learns From Experiencecited by 280
  11. 2025LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Modelscited by 278
  12. 2023Task Arithmetic in the Tangent Space: Improved Editing of Pre-Trained Modelscited by 218
  13. 2023Scaling Robot Learning with Semantically Imagined Experiencecited by 175
  14. 2025LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorizationcited by 166
  15. 2024Behavior Generation with Latent Actionscited by 159
  16. 2024Task Singular Vectors: Reducing Task Interference in Model Mergingcited by 143
  17. 2024Model merging with SVD to tie the Knotscited by 118
  18. 2025No Task Left Behind: Isotropic Model Merging with Common and Task-Specific Subspacescited by 90
  19. 2024Model Stock: All we need is just a few fine-tuned modelscited by 87
  20. 2023Decomposing the Generalization Gap in Imitation Learning for Visual Robotic Manipulationcited by 78
  21. 2024OPEN TEACH: A Versatile Teleoperation System for Robotic Manipulationcited by 67
  22. 2025Whoever Started the Interference Should End It: Guiding Data-Free Model Merging via Task Vectorscited by 59
  23. 2025Evo-0: Vision-Language-Action Model with Implicit Spatial Understandingcited by 55
  24. 2023Chat Vector: A Simple Approach to Equip LLMs with Instruction Following and Model Alignment in New Languagescited by 52
  25. 2024RoVi-Aug: Robot and Viewpoint Augmentation for Cross-Embodiment Robot Learningcited by 46
  26. 2025Novel Demonstration Generation with Gaussian Splatting Enables Robust One-Shot Manipulationcited by 45
  27. 2024AdaMergeX: Cross-Lingual Transfer with Large Language Models via Adaptive Adapter Mergingcited by 34
  28. 2023Language and Task Arithmetic with Parameter-Efficient Layers for Zero-Shot Summarizationcited by 28
  29. 2025Accurate and Efficient Low-Rank Model Merging in Core Spacecited by 28
  30. 2024ReVLA: Reverting Visual Domain Limitation of Robotic Foundation Modelscited by 26
  31. 2024On the Emergence of Cross-Task Linearity in the Pretraining-Finetuning Paradigmcited by 25
  32. 2025DataMIL: Selecting Data for Robot Imitation Learning with Datamodelscited by 25
  33. 2025GeoAware-VLA: Implicit Geometry Aware Vision-Language-Action Modelcited by 24
  34. 2025Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Mergingcited by 23
  35. 2024SoMA: Singular Value Decomposed Minor Components Adaptation for Domain Generalizable Representation Learningcited by 22
  36. 2025CAT Merging: A Training-Free Approach for Resolving Conflicts in Model Mergingcited by 22
  37. 2025Robust Finetuning of Vision-Language-Action Robot Policies via Parameter Mergingcited by 19
  38. 2025Adapt3R: Adaptive 3D Scene Representation for Domain Transfer in Imitation Learningcited by 16

50 most cited of 65 All 65 (JSON)

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Generalizable VLA Finetuning via Representation Anchoring and Language-Action Alignment
  2. 2025RynnVLA-002: A Unified Vision-Language-Action and World Model
  3. 2025VLA-0: Building State-of-the-Art VLAs with Zero Modification
  4. 2026PriorVLA: Prior-Preserving Adaptation for Vision-Language-Action Models
  5. 2025VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Model
  6. 2026VLA-REPLICA: A Low-Cost, Reproducible Benchmark for Real-World Evaluation of Vision-Language-Action Models
  7. 2025MoS-VLA: A Vision-Language-Action Model with One-Shot Skill Adaptation
  8. 2026CLAP: Direct VLM-to-VLA Adaptation via Language-Action Grounding