Metraipolis › archive › dc_8fc0d7bb563e

Seeing Parts, Reasoning about Worlds: Visual Inference under Partial Observation — arXiv:2609.32456

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.32456
License (as archived): CC0-1.0
Archived: 2026-10-07T21:00:14.618Z
SHA-256 of the archived text: fba07a643ad8e2789e3d3ec8dba45e47ba5389718cee7b176edeec611a12cdb8
Stable URL of this page: https://metraipolis.com/documento/dc_8fc0d7bb563e

How to cite

Seeing Parts, Reasoning about Worlds: Visual Inference under Partial Observation — arXiv:2609.32456 — arXiv (en), https://arxiv.org/abs/2609.32456 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8fc0d7bb563e (Metraipolis archive dc_8fc0d7bb563e, sha256 fba07a643ad8e2789e3d3ec8dba45e47ba5389718cee7b176edeec611a12cdb8)

BibTeX

@misc{dc8fc0d7bb563e,
  title = {Seeing Parts, Reasoning about Worlds: Visual Inference under Partial Observation — arXiv:2609.32456},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.32456},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8fc0d7bb563e; sha256 fba07a643ad8e2789e3d3ec8dba45e47ba5389718cee7b176edeec611a12cdb8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.32456. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

World modeling under partial observation requires reasoning about the complete worlds that remain compatible with limited visual evidence. Occluded objects and unseen regions can leave several world states possible; additional views can exclude alternatives and strengthen the conclusions supported by the observations. We introduce WorldScope to study this process through possible-world semantics, evidence-grounded data, and learned visual representations. WorldScope-1.2M provides 1.2 million English question-answer pairs spanning eight world properties, ten task interfaces, and three observation protocols. Its answers encode confirmed facts, supported bounds, and unresolved possibilities. Complementary supervision comprises 4,800 certified counterworld groups with equivalent base observations and different hidden object configurations and query answers. These groups provide physical witnesses of ambiguity and training-only labels for world compatibility and view-induced exclusions. We propose WorldFlow, which composes cross-view entity evidence and support-surface coverage into an image-subset evidence lattice. Counterworld compatibility and transition objectives train subset representations to reflect how new observations constrain possible worlds. A shared answer generator uses these representations to predict the strongest supported conclusion. WorldScope-Bench evaluates claim judgments and evidence-dependent conclusions as views are selected, combined, removed, or ordered. On its 5,000-question test set, WorldFlow reaches 64.34% exact accuracy, improving over the same backbone trained on QA alone by 24.88 percentage points. It retains 50.43% accuracy on the 3,000 questions from structure-disjoint scenes.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.32456 — 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 2026-10-07T20:29:02.343Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).