Metraipolis › archive › dc_514e2d82f7dc

Roz: A Taste Test for World Model Latents — arXiv:2611.00612

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00612
License (as archived)
CC0-1.0
Archived
2026-11-03T11:24:52.051Z
SHA-256 of the archived text
4cef9d14a695ea1fc87faa893c731c6cf7a542a06688e054e5bc400d72b9bf44
Stable URL of this page
https://metraipolis.com/documento/dc_514e2d82f7dc

How to cite

Roz: A Taste Test for World Model Latents — arXiv:2611.00612 — arXiv (en), https://arxiv.org/abs/2611.00612 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_514e2d82f7dc (Metraipolis archive dc_514e2d82f7dc, sha256 4cef9d14a695ea1fc87faa893c731c6cf7a542a06688e054e5bc400d72b9bf44)
BibTeX
@misc{dc514e2d82f7dc,
  title = {Roz: A Taste Test for World Model Latents — arXiv:2611.00612},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00612},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_514e2d82f7dc; sha256 4cef9d14a695ea1fc87faa893c731c6cf7a542a06688e054e5bc400d72b9bf44},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00612. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

State of the art latent dynamics world models achieve impressive performance for robot planning. But do they actually encode the physics of the world? In this paper, we show that popular world models today achieve strong results on tasks that reduce to image similarity but forget key details about objects in the scene like mass, deformation, friction, and bounciness. We achieve this by defining precisely what ought to be encoded in such a world model and introducing two new evaluation methods that can directly tease physical properties and easily scale to new ones. Our evaluation, which we call Roz, includes 10 contrastive tasks and 3 continuous tasks, totaling 1734 individual scenarios. We validate our eval on control tasks. On our own model, AnonWM, we find that performance roughly improves with training steps, while not necessarily being correlated with reconstruction MSE. We argue current model failures are a symptom of evaluation strategies (probes and planners) that are opaque, model-specific, and high-compute. Roz requires only a model that can ingest still frames or video, a way to extract latents, and a distance function between them, and it can be run in one pass. We hope future evaluations will adopt Roz's evaluation techniques. More broadly, we hope Roz can be a case study for a future where benchmarks allow rapid iteration and comparison across models.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.00612 — 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 2026-11-03T11:02:37.775Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).