Metraipolis › archive › dc_e4b196b2f477

VLA-ACL: Action-Consistent Visual Token Pruning for Efficient Vision-Language-Action Models — arXiv:2610.08133

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.08133
License (as archived)
CC0-1.0
Archived
2026-10-07T13:07:59.156Z
SHA-256 of the archived text
5e3a6f75e1a35281de0d63d527ea848a1b7b0551d13f5f4366010964bbd80845
Stable URL of this page
https://metraipolis.com/documento/dc_e4b196b2f477

How to cite

VLA-ACL: Action-Consistent Visual Token Pruning for Efficient Vision-Language-Action Models — arXiv:2610.08133 — arXiv (en), https://arxiv.org/abs/2610.08133 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e4b196b2f477 (Metraipolis archive dc_e4b196b2f477, sha256 5e3a6f75e1a35281de0d63d527ea848a1b7b0551d13f5f4366010964bbd80845)
BibTeX
@misc{dce4b196b2f477,
  title = {VLA-ACL: Action-Consistent Visual Token Pruning for Efficient Vision-Language-Action Models — arXiv:2610.08133},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.08133},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e4b196b2f477; sha256 5e3a6f75e1a35281de0d63d527ea848a1b7b0551d13f5f4366010964bbd80845},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08133. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-Language-Action (VLA) models achieve strong robotic manipulation performance but incur high computational costs from processing long token sequences at every control step, limiting real-time deployment. Visual token pruning offers a direct solution, as visual patches dominate the input sequence and contain considerable redundancy. Existing approaches, however, either rely on indirect training-free heuristics, such as attention scores and motion thresholds, or require costly fine-tuning of the base VLA model. We introduce VLA-ACL (Action Consistency Learning), which learns a lightweight visual token pruning policy through action-level supervision while keeping the base VLA model entirely frozen. The training objective encourages actions produced from pruned visual contexts to remain consistent with the full-context teacher, with ground-truth actions as auxiliary supervision. This directly ties token selection to its effect on the downstream control output. Experiments on LIBERO and real-world manipulation tasks show that VLA-ACL prunes up to 87.5% of visual tokens while retaining competitive performance, reduces computation by up to 75%, and achieves a 1.5x inference speedup. These results establish a stronger performance-efficiency trade-off than existing frozen-VLA pruning methods and demonstrate the value of action-level supervision for visual token selection. Code is available at this https URL .

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.08133 — 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 2026-10-07T12:37:33.045Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).