Skip to content
AI News HubLIVE
Original source3 min read

RageVLA: Augmentation-Trained Perception and Action for State-of-the-Art Robustness on LIBERO

Summary

RageVLA argues that VLA models saturate on LIBERO yet collapse on LIBERO-PRO because what to do stays tied to training-time object positions, surrounding scenes, and instruction wording. The authors cut these ties in the data (general rules that move and rotate every object and fixture, port tasks to other environments and surfaces, and paraphrase every instruction) and in the policy (a perception engine on frozen DINOv3 features hands the action model a decoupled view: a list of objects with positions and a compact gripper-focus code alongside TinyCLIP image and instruction embeddings, with a from-scratch flow-matching diffusion transformer acting on it). The 0.21B-parameter model scores 0.72 on LIBERO-PRO versus 0.53 for the best published model, with a larger variant at 0.73. LIBERO-Ob…

SourcearXiv RoboticsAuthor: Amrit Lal Singh, Smarak Kanjilal
RageVLA: Augmentation-Trained Perception and Action for State-of-the-Art Robustness on LIBERO
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 17 Nov 2026]

Title:RageVLA: Augmentation-Trained Perception and Action for State-of-the-Art Robustness on LIBERO

View a PDF of the paper titled RageVLA: Augmentation-Trained Perception and Action for State-of-the-Art Robustness on LIBERO, by Amrit Lal Singh and 1 other authors

View PDF HTML (experimental)

Abstract:Vision-language-action (VLA) models look saturated on LIBERO, yet most of them fall apart on LIBERO-PRO as soon as an object moves or the instruction asks for something new. We argue that what is missing is not scale but decoupling.

In these policies, what to do is tied to where things were during training, to the scene around them and to the wording of the instruction. We cut these ties in two places. In the data, LIBERO demonstrations are re-arranged by general rules: every object and fixture is moved and rotated, tasks are carried into other environments and onto other surfaces, and every instruction is paraphrased. In the policy, a perception engine on frozen DINOv3 features, trained on this augmented data, hands the action model the scene in decoupled form: a list of objects with their positions, and a compact code of what the gripper is looking at, next to image and instruction embeddings from TinyCLIP. A flow-matching diffusion transformer, trained from scratch, acts on this view.

We call this combination of Robust Augmentation and Grounded Encoding RageVLA. With 0.21 billion parameters, RageVLA sets a new state of the art on LIBERO-PRO among VLA policies, scoring 0.72 against 0.53 for the best published model, and a larger variant reaches 0.73.

Two further tests on LIBERO-Object show that the policy acts on what it perceives. When the object positions it is given are shuffled while the camera images stay correct, its success falls from 0.965 to 0.215. When the target object is displaced by 28 cm, it still succeeds in 0.865 of the episodes, where $\pi_{0.5}$, run through the same harness, succeeds in 0.105.

Comments: 10 pages, 10 figures, 9 tables

Subjects:

Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)

Cite as: arXiv:2611.23517 [cs.RO]

(or arXiv:2611.23517v1 [cs.RO] for this version)

https://doi.org/10.48550/arXiv.2611.23517

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Amrit Lal Singh [view email] [v1] Tue, 17 Nov 2026 21:07:47 UTC (662 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled RageVLA: Augmentation-Trained Perception and Action for State-of-the-Art Robustness on LIBERO, by Amrit Lal Singh and 1 other authors

View PDF

HTML (experimental)

TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.RO

new | recent | 2026-11

Change to browse by:

cs cs.CV

References & Citations

SciX

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • RageVLA reaches 0.72 on LIBERO-PRO with 0.21B parameters, beating the best published model's 0.53, while a larger variant hits 0.73
  • The fix is decoupling rather than scale: rule-based data rearrangement plus a perception engine that gives the action model a decoupled scene view
  • Shuffling the provided object positions while keeping camera images correct drops success from 0.965 to 0.215, showing the policy relies on its perceived positions
  • With the target displaced by 28 cm, RageVLA still succeeds in 0.865 of episodes versus 0.105 for π0.5 in the same harness

Highlights and analysis are generated automatically and may contain errors. Check the original source.