Metraipolis › archive › dc_1fc46356495e

Beyond Behaviour: A Stimulus-Response Safety Evaluation of Vision-Language-Action Models — arXiv:2610.34837

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.34837
License (as archived)
CC0-1.0
Archived
2026-10-27T07:58:12.153Z
SHA-256 of the archived text
2998693d526a213f29fcd1bc5371d3016401715571bec0e26291f30e72e978f6
Stable URL of this page
https://metraipolis.com/documento/dc_1fc46356495e

How to cite

Beyond Behaviour: A Stimulus-Response Safety Evaluation of Vision-Language-Action Models — arXiv:2610.34837 — arXiv (en), https://arxiv.org/abs/2610.34837 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1fc46356495e (Metraipolis archive dc_1fc46356495e, sha256 2998693d526a213f29fcd1bc5371d3016401715571bec0e26291f30e72e978f6)
BibTeX
@misc{dc1fc46356495e,
  title = {Beyond Behaviour: A Stimulus-Response Safety Evaluation of Vision-Language-Action Models — arXiv:2610.34837},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.34837},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1fc46356495e; sha256 2998693d526a213f29fcd1bc5371d3016401715571bec0e26291f30e72e978f6},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.34837. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Humans exhibit biological sensitivity to safety-critical stimuli that can trigger defensive action. Across three models and two video benchmarks, we investigate whether similar responses emerge in vision-language-action models by examining their internal representations and action tendencies in response to visible hazards and harmful consequences. Our findings suggest that task execution may remain the default response even when danger is internally registered.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.34837 — 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 2026-10-27T07:55:30.719Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).