Metraipolis › archive › dc_7e67d8d7756d

Training Object Permanence in World Models — arXiv:2609.28654

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.28654
License (as archived)
CC0-1.0
Archived
2026-10-08T04:25:38.254Z
SHA-256 of the archived text
eb16417d1262fa8f97d4749cfdb62aa09dfbc9ad40acb68b8e8bacd20d19fa35
Stable URL of this page
https://metraipolis.com/documento/dc_7e67d8d7756d

How to cite

Training Object Permanence in World Models — arXiv:2609.28654 — arXiv (en), https://arxiv.org/abs/2609.28654 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7e67d8d7756d (Metraipolis archive dc_7e67d8d7756d, sha256 eb16417d1262fa8f97d4749cfdb62aa09dfbc9ad40acb68b8e8bacd20d19fa35)
BibTeX
@misc{dc7e67d8d7756d,
  title = {Training Object Permanence in World Models — arXiv:2609.28654},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.28654},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7e67d8d7756d; sha256 eb16417d1262fa8f97d4749cfdb62aa09dfbc9ad40acb68b8e8bacd20d19fa35},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28654. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Object permanence and solidity are hallmarks of human cognitive priors. Recent studies show that video generation models, a paradigmatic class of current world models, have begun to show emerged reasoning abilities, making them ideal candidates for building human-like physical intelligence. Do video models have emerged object permanence in them? If not, could we train them with a core-cognition inspired dataset? We introduce WROP (World Reasoning with Object Permanence), a data infrastructure of 150 hand-designed cognitive science inspired tasks, divided into six cognitive categories. We build Blender generators that randomize speed, lighting, camera angle, and other nuisance parameters while preserving each task's cognitive structure, yielding 10,000+ samples per task. We release a 1.5M-sample training corpus and a 300-question exam. On this exam we evaluate 14 video models: 3 reference-to-video, 7 edit, and 4 continuation, among which PWM-WROP, our 16B world model. In a blind pairwise Elo study, PWM-WROP ranks first among continuation models and third overall, behind only a statistical tie between two reference-to-video models. We release the data, exam, model answers, scores, weights, and PWM, our native-PyTorch training stack on AWS Trainium2.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-25 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.28654 — 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 2026-10-08T03:50:57.664Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).