Metraipolis › archive › dc_1efdf2361516

Sprout: Building Dynamic Memory While Reasoning for Agentic Video Understanding — arXiv:2609.35497

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.35497
License (as archived)
CC0-1.0
Archived
2026-10-07T23:54:49.564Z
SHA-256 of the archived text
71129456732c5d3bac274995ca76e544b9aed76c4f5ec51855f6acd6c09945df
Stable URL of this page
https://metraipolis.com/documento/dc_1efdf2361516

How to cite

Sprout: Building Dynamic Memory While Reasoning for Agentic Video Understanding — arXiv:2609.35497 — arXiv (en), https://arxiv.org/abs/2609.35497 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1efdf2361516 (Metraipolis archive dc_1efdf2361516, sha256 71129456732c5d3bac274995ca76e544b9aed76c4f5ec51855f6acd6c09945df)
BibTeX
@misc{dc1efdf2361516,
  title = {Sprout: Building Dynamic Memory While Reasoning for Agentic Video Understanding — arXiv:2609.35497},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.35497},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1efdf2361516; sha256 71129456732c5d3bac274995ca76e544b9aed76c4f5ec51855f6acd6c09945df},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35497. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long video understanding relies on video memory to overcome the context limits of multimodal large language models. Existing methods follow a build-then-reasoning pipeline: memory is built offline for the entire video, then reasoned over as a static source. In practice a long video is shared by several questions, and this pipeline is costly at both ends: with few questions, building memory for the whole video costs far more than answering them; with many questions, the memory is never updated, so what is learned while answering questions is lost to the next question. To alleviate these, we introduce Sprout, an agentic framework that builds memory while reasoning: a temporal tree that sprouts detailed nodes as questions are answered. The agent watches the video segment by segment at a low frame rate, stopping when the current question can be answered, remembers each segment as a coarse node of the tree, and revisits key intervals at a higher frame rate to refine the tree with the recovered details. Once a segment is recorded as text, its video input is removed from the context history, while the original video remains reachable through the video tools. The memory tree and prior question--answer records persist across questions, so the memory is online and dynamic: built from the first question onward and updated by every question thereafter. We find that replacing accumulated video inputs with textual memory substantially reduces context usage while maintaining accuracy, with slight improvements in some settings. Across benchmarks on three models, Sprout achieves competitive or improved accuracy relative to representative offline memory methods, with no upfront construction stage and lower context cost per question.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-29 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.35497 — 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 2026-10-07T23:29:27.581Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).