Metraipolis › archive › dc_7d7f6d1db653

Do VLAs Really Follow Instructions? Diagnosing How-Level Instruction Following in Vision-Language-Action Models — arXiv:2610.21251

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.21251
License (as archived)
CC0-1.0
Archived
2026-10-16T10:35:14.532Z
SHA-256 of the archived text
d5721d6c152b469a85728dea2206b5d20c13169b99c20a7f5d7ad278ed80093a
Stable URL of this page
https://metraipolis.com/documento/dc_7d7f6d1db653

How to cite

Do VLAs Really Follow Instructions? Diagnosing How-Level Instruction Following in Vision-Language-Action Models — arXiv:2610.21251 — arXiv (en), https://arxiv.org/abs/2610.21251 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7d7f6d1db653 (Metraipolis archive dc_7d7f6d1db653, sha256 d5721d6c152b469a85728dea2206b5d20c13169b99c20a7f5d7ad278ed80093a)
BibTeX
@misc{dc7d7f6d1db653,
  title = {Do VLAs Really Follow Instructions? Diagnosing How-Level Instruction Following in Vision-Language-Action Models — arXiv:2610.21251},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.21251},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7d7f6d1db653; sha256 d5721d6c152b469a85728dea2206b5d20c13169b99c20a7f5d7ad278ed80093a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.21251. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-language-action (VLA) models are expected to follow instructions like their language backbones, yet prior work has focused on What-level instructions that specify the task rather than How-level instructions that specify how to carry it out. Through our diagnoses, we find that How-level instructions do not help current VLAs and lower task success. We hypothesize that pairing each training episode with a single instruction limits How-level instruction following, and show that training with segment-level enriched instructions makes actions respond to such instructions without consistently raising task success. To measure How-level instruction following beyond standard benchmarks, we introduce SIMPLER-IN, where success requires conditions given only in the instruction. It shows that the gained responsiveness does not extend to demanding or out-of-distribution How-level instructions. How-level instruction following thus remains an open challenge.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2610.21251 — 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 2026-10-16T08:52:02.645Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).