Metraipolis › archive › dc_f281e24bc221

Are Vision-Language-Action Models Robust to One-Step Observation Perturbations? — arXiv:2609.32550

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.32550
License (as archived)
CC0-1.0
Archived
2026-10-07T20:55:42.903Z
SHA-256 of the archived text
99a4e6bb0f1cbcf4c1fb0e826e0db3d774da8772de400f15b0e087097ce7da5e
Stable URL of this page
https://metraipolis.com/documento/dc_f281e24bc221

How to cite

Are Vision-Language-Action Models Robust to One-Step Observation Perturbations? — arXiv:2609.32550 — arXiv (en), https://arxiv.org/abs/2609.32550 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f281e24bc221 (Metraipolis archive dc_f281e24bc221, sha256 99a4e6bb0f1cbcf4c1fb0e826e0db3d774da8772de400f15b0e087097ce7da5e)
BibTeX
@misc{dcf281e24bc221,
  title = {Are Vision-Language-Action Models Robust to One-Step Observation Perturbations? — arXiv:2609.32550},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.32550},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f281e24bc221; sha256 99a4e6bb0f1cbcf4c1fb0e826e0db3d774da8772de400f15b0e087097ce7da5e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.32550. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Understanding the safety risks of vision-language-action (VLA) models is essential for their deployment in the physical world. Existing safety research has mainly considered persistent perturbations that are applied continuously to observations throughout an episode. However, momentary observation corruption, in which observations are severely perturbed only briefly within an episode, remains an underexplored safety threat. To address this gap, this work investigates robustness to one-step perturbations applied at a single time step per episode. Our experiments reveal that these perturbations substantially degrade VLA performance and that their impact depends on the action chunk execution length. Based on them, we propose CARE, which dynamically selects the execution length based on consistency with the previously predicted action chunk. CARE improves robustness with low computational overhead while preserving clean performance by selecting shorter execution lengths only under perturbations.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Machine Learning (cs.LG); Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2609.32550 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T20:29:00.870Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).