Metraipolis › archive › dc_856a5a4753ff

Quantize by Doing: Task-Guided Mixed-Precision Quantization of Vision-Language-Action Models — arXiv:2610.38661

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.38661
License (as archived)
CC0-1.0
Archived
2026-11-01T02:44:16.365Z
SHA-256 of the archived text
0d630c1e3a05dcb3ad0f0d28704173bffc78eebf0ae8428f261342a19f3e194b
Stable URL of this page
https://metraipolis.com/documento/dc_856a5a4753ff

How to cite

Quantize by Doing: Task-Guided Mixed-Precision Quantization of Vision-Language-Action Models — arXiv:2610.38661 — arXiv (en), https://arxiv.org/abs/2610.38661 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_856a5a4753ff (Metraipolis archive dc_856a5a4753ff, sha256 0d630c1e3a05dcb3ad0f0d28704173bffc78eebf0ae8428f261342a19f3e194b)
BibTeX
@misc{dc856a5a4753ff,
  title = {Quantize by Doing: Task-Guided Mixed-Precision Quantization of Vision-Language-Action Models — arXiv:2610.38661},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.38661},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_856a5a4753ff; sha256 0d630c1e3a05dcb3ad0f0d28704173bffc78eebf0ae8428f261342a19f3e194b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.38661. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Post-training quantization (PTQ) reduces the memory footprint of vision-language-action (VLA) models, but lowering bit widths while maintaining task performance remains challenging. Existing work has used sensitivity estimates from calibration data to assign bit widths to different model components. However, using task success directly to guide mixed-precision allocation is underexplored because closed-loop evaluations are costly and the effect of quantizing one component depends on the precision of other components. To this end, we propose \textbf{QbD-VLA}, a task-guided mixed-precision quantization method that uses task success from short closed-loop evaluations to assign weight and activation bit widths while accounting for interactions between model components. Experiments covering four VLA models and all four LIBERO suites show strong task performance at average precisions of approximately 4 bits for weights and fewer than 6 bits for activations. On SIMPLER, the gain over uniform quantization is small. On a Franka Research 3 robot arm, simulation-selected allocations show no detected success-rate loss on X-VLA but fail on $\pi_{0.5}$. These findings support using task success to guide mixed-precision VLA quantization, while highlighting the need for real-world evaluation before deployment. Code is available at \href{ this https URL }{\texttt{ this http URL }}.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.RO, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2610.38661 — 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 2026-11-01T02:16:58.635Z dalla pagina https://arxiv.org/list/cs.RO/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).