Metraipolis › archive › dc_eeaf60b4cdab

VidHarness: Evolving Agent Harnesses for Cost-Efficient Long Video Understanding — arXiv:2609.38413

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.38413
License (as archived)
CC0-1.0
Archived
2026-10-07T14:45:16.210Z
SHA-256 of the archived text
6fc0f33da182bfcee5ed5d0f266f97e8cc7a146c70946b3d80ba9375a9613036
Stable URL of this page
https://metraipolis.com/documento/dc_eeaf60b4cdab

How to cite

VidHarness: Evolving Agent Harnesses for Cost-Efficient Long Video Understanding — arXiv:2609.38413 — arXiv (en), https://arxiv.org/abs/2609.38413 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eeaf60b4cdab (Metraipolis archive dc_eeaf60b4cdab, sha256 6fc0f33da182bfcee5ed5d0f266f97e8cc7a146c70946b3d80ba9375a9613036)
BibTeX
@misc{dceeaf60b4cdab,
  title = {VidHarness: Evolving Agent Harnesses for Cost-Efficient Long Video Understanding — arXiv:2609.38413},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.38413},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eeaf60b4cdab; sha256 6fc0f33da182bfcee5ed5d0f266f97e8cc7a146c70946b3d80ba9375a9613036},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38413. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-language models (VLMs) can answer questions about hour-long videos, but processing every frame is prohibitively expensive, even though the evidence for a question usually spans only a few seconds. Video agents, i.e., harness programs wrapped around a frozen VLM, address this by observing the video selectively, yet existing harnesses are hand-crafted by experts through slow build-and-test cycles. We propose VidHarness, a framework that automates harness design for cost-efficient long video understanding, in which a harness proposer iteratively evolves harnesses based on execution feedback from an evolution environment. To escape the local optima of greedy refinement, we organize the evolution as Monte Carlo tree search (MCTS), and to reduce the evaluation cost, we integrate uncertainty-aware multi-fidelity validation, which screens new harnesses on a few questions and promotes only the promising ones. Since the best harness varies with the frame budget, we further introduce a mixture-of-harness that routes each question to a harness specialized for its budget. VidHarness sets new state-of-the-art results on LongVideoBench, Video-MME, and Video-Holmes, outperforms the strongest hand-crafted video agent by up to $11.2$ points, and generalizes to the knowledge-intensive benchmarks Video-MMMU and MMVU with fewer than half of the frames of uniform sampling.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.38413 — 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 2026-10-07T14:19:00.554Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).