Metraipolis › archive › dc_a01224bbf448

Not All Error Yields to Scale: Where Scaling Stops in Vision-Language Inference — arXiv:2610.01640

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.01640
License (as archived)
CC0-1.0
Archived
2026-10-06T12:14:38.585Z
SHA-256 of the archived text
d6dbfa72ffa02f3a3247983066fee42f7c4f59022084951152e25c57348aa3b2
Stable URL of this page
https://metraipolis.com/documento/dc_a01224bbf448

How to cite

Not All Error Yields to Scale: Where Scaling Stops in Vision-Language Inference — arXiv:2610.01640 — arXiv (en), https://arxiv.org/abs/2610.01640 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a01224bbf448 (Metraipolis archive dc_a01224bbf448, sha256 d6dbfa72ffa02f3a3247983066fee42f7c4f59022084951152e25c57348aa3b2)
BibTeX
@misc{dca01224bbf448,
  title = {Not All Error Yields to Scale: Where Scaling Stops in Vision-Language Inference — arXiv:2610.01640},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.01640},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a01224bbf448; sha256 d6dbfa72ffa02f3a3247983066fee42f7c4f59022084951152e25c57348aa3b2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.01640. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-language models (VLMs) face a fixed-budget trade-off between processing more visual information for fine-grained perception and using a larger language backbone for complex reasoning. Existing studies do not tell us which combination of backbone size and input resolution to deploy, especially in high-resolution deployments. To address this gap, we propose the Separable Law that describes how VLM performance changes with language backbone size and visual token count. We fit the law to measurements from 26 InternVL and QwenVL models, with language backbone sizes from 1B to 72B, on four high-resolution benchmarks with image sizes from 224 pixels to 8K. We find that the questions responding to scaling can be predicted from the skill they require, while a substantial fraction never responds at all. We also find that the two model families gain similarly from a larger backbone, while their gains from more visual tokens differ sharply. Combined with a cost law, the Separable Law gives a closed-form rule for allocating compute between backbone size and visual tokens. When deployment is limited to available configurations, the law identifies model and image sizes that perform close to the best feasible choice under the same budget. We hope our work offers a principled way to decide how much a model should be allowed to see at high resolution, given what it must reason about.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-02 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.01640 — 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 2026-10-06T10:21:28.205Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).