Metraipolis › archive › dc_66958abf366e

Visual Orchestration Tax in Agentic VLM Pipelines: Auditing and Certifying Visual Evidence Reuse — arXiv:2610.08170

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.08170
License (as archived)
CC0-1.0
Archived
2026-10-07T11:17:12.837Z
SHA-256 of the archived text
03e55a54a2c2e8419fb705ccdd9622950ef75f85fdf9d7021f52d9f1b98b507a
Stable URL of this page
https://metraipolis.com/documento/dc_66958abf366e

How to cite

Visual Orchestration Tax in Agentic VLM Pipelines: Auditing and Certifying Visual Evidence Reuse — arXiv:2610.08170 — arXiv (en), https://arxiv.org/abs/2610.08170 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_66958abf366e (Metraipolis archive dc_66958abf366e, sha256 03e55a54a2c2e8419fb705ccdd9622950ef75f85fdf9d7021f52d9f1b98b507a)
BibTeX
@misc{dc66958abf366e,
  title = {Visual Orchestration Tax in Agentic VLM Pipelines: Auditing and Certifying Visual Evidence Reuse — arXiv:2610.08170},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.08170},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_66958abf366e; sha256 03e55a54a2c2e8419fb705ccdd9622950ef75f85fdf9d7021f52d9f1b98b507a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08170. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agentic VLM pipelines increasingly pass the same static visual evidence through multiple specialist agents and tools. This design creates an orchestration-level redundancy mode: semantically unchanged images are repeatedly reconstructed as image-conditioned requests at the VLM API boundary. We call this phenomenon visual orchestration tax and develop a measurement-to-certification framework for visual evidence reuse in agentic VLM pipelines. The audit side defines $\mathrm{M1}_{\mathrm{trace}}$ to count raw visual-evidence touches and M2 to measure structural touch redundancy, with query-level distributions, bootstrap confidence intervals, and paired quality tests. Across SeeingEye and MAMMQA on chart, document, general-VQA, and multi-modal-QA tasks, audits reveal 66.8-75.6% visual-evidence touch redundancy, and every audited query exceeds the predefined gate. The certification side introduces SharedVisCache, a contract-aware evidence reuse hook keyed by image content, preprocessing fingerprint, and encoder assumptions. On SeeingEye, contract validation certifies 75.0-75.5% repeated touches as reusable while preserving 350/350 output strings and $\Delta\mathrm{M5}{=}0$. At the physical layer, certified hits reduce $F_{\mathrm{vision}}$ from 800 to 200 in ChartQA-200 trace replay and from 200 to 50 inside live SeeingEye translator-stage physical integration, preserving 800/800 replay strings and 200/200 integrated call outputs. The results position visual reuse as a measurable, behavior-preserving property of agent orchestration and define an agent-layer contract that makes backend prefix or token reuse semantically interpretable.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno).
Categorie arXiv: Multiagent Systems (cs.MA) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.08170 — 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 2026-10-07T11:15:09.461Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).