Metraipolis › archive › dc_ae42e85c7d4f

From Intent to Evidence: Policy-Steered Multi-Strategy Retrieval for Long-Video Agents — arXiv:2608.31005

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.31005
License (as archived)
CC0-1.0
Archived
2026-10-17T09:59:46.674Z
SHA-256 of the archived text
345d59fbb401908edf0df41c7d765ed3ca50c5a5ceca4aef1b5f30eb82ff35af
Stable URL of this page
https://metraipolis.com/documento/dc_ae42e85c7d4f

How to cite

From Intent to Evidence: Policy-Steered Multi-Strategy Retrieval for Long-Video Agents — arXiv:2608.31005 — arXiv (en), https://arxiv.org/abs/2608.31005 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ae42e85c7d4f (Metraipolis archive dc_ae42e85c7d4f, sha256 345d59fbb401908edf0df41c7d765ed3ca50c5a5ceca4aef1b5f30eb82ff35af)
BibTeX
@misc{dcae42e85c7d4f,
  title = {From Intent to Evidence: Policy-Steered Multi-Strategy Retrieval for Long-Video Agents — arXiv:2608.31005},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.31005},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ae42e85c7d4f; sha256 345d59fbb401908edf0df41c7d765ed3ca50c5a5ceca4aef1b5f30eb82ff35af},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.31005. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Existing long-video agents acquire evidence through one uniform behavior, ignoring whether the required evidence is concentrated, requires broad occurrence coverage, or must discriminate competing hypotheses---which can cause failure before substantive reasoning begins. Prescribing a fine-grained solution procedure for every question is not a satisfactory remedy, as it restricts autonomous exploration. We propose VESTA, a training-free long-video agent organized as a route-conditioned acquire--verify--consolidate loop. Before exploration, an intent router infers an evidence-acquisition policy---focused, recall, or contrastive retrieval over a shared visual--speech scene index---together with an evidence-accounting policy that configures the evidence view maintained during exploration. Policy-steered retrieval yields provisional references that multimodal evidence operations convert into observations, while the Reasoner remains free to verify them, re-query using intermediate findings, or inspect regions outside the retrieved set. A temporal evidence ledger consolidates observations into an adaptive, compressed view of temporal location, provenance, coverage, conflicts, verification outcomes, and hypothesis support, exposing missing and unresolved evidence to guide subsequent acquisition; finalization prioritizes verified observations. On Video-MME-v2, VESTA improves average accuracy by 2.7 points over VideoARM and gains across all six reported metrics. On LongVideoBench, EgoSchema, and LVBench under shared query-time models, it improves by 6.9 points on the LongVideoBench long subset and 1.5 on LVBench, and matches VideoARM on EgoSchema.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-01 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.31005 — 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 2026-10-17T09:28:48.208Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).