Metraipolis › archive › dc_c4094de80d7f

DynaPix: Can Vision-Language Models Identify the Exact Future? — arXiv:2608.05505

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.05505
License (as archived): CC0-1.0
Archived: 2026-10-31T10:35:57.187Z
SHA-256 of the archived text: 8fc7be49fd1120cc129d8e09f3de68c4d23a59b9a1f4ab0afa47fbd619c749b4
Stable URL of this page: https://metraipolis.com/documento/dc_c4094de80d7f

How to cite

DynaPix: Can Vision-Language Models Identify the Exact Future? — arXiv:2608.05505 — arXiv (en), https://arxiv.org/abs/2608.05505 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c4094de80d7f (Metraipolis archive dc_c4094de80d7f, sha256 8fc7be49fd1120cc129d8e09f3de68c4d23a59b9a1f4ab0afa47fbd619c749b4)

BibTeX

@misc{dcc4094de80d7f,
  title = {DynaPix: Can Vision-Language Models Identify the Exact Future? — arXiv:2608.05505},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.05505},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c4094de80d7f; sha256 8fc7be49fd1120cc129d8e09f3de68c4d23a59b9a1f4ab0afa47fbd619c749b4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.05505. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Acting in a physical scene requires knowing its real later state, not a plausible one. Current evaluations often accept words or a realistic-looking image, so the predicted state is never checked against the true one. We introduce DynaPix (Dynamic Pixels), a benchmark that makes prediction checkable. Given a video clip that stops before a key event and a question about a later moment, a model must pick the true future image from close candidates or a large gallery. The scenes come from a physics simulator, so the correct image and its time are known exactly and the wrong options are deliberately similar. Models often succeed when a visible event marks the target moment, but are near chance when only elapsed time marks it. Gallery search is harder still, as the true image rarely ranks first. People handle the elapsed-time items well, so the difficulty lies with the models, not the questions. Training on scene accounts drawn from the simulator's true record, not a teacher's guess, repairs much of this but not the longer elapsed time case. DynaPix thus exposes a temporal-anchoring gap: models attach a prediction to an event far better than to time itself.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-07 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.05505 — 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 2026-10-31T10:23:30.978Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-07?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).