Metraipolis › archive › dc_9075f6931ffc
Modeling What Changes: Sparse, Residual World Models for Object-Centric Manipulation — arXiv:2609.02046
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.02046
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T03:36:58.979Z
- SHA-256 of the archived text
35c8cc746155a2eb18aeb3ed041d610238532cb06d7b7d39f0598fbbf5bb4dd6- Stable URL of this page
https://metraipolis.com/documento/dc_9075f6931ffc
How to cite
Modeling What Changes: Sparse, Residual World Models for Object-Centric Manipulation — arXiv:2609.02046 — arXiv (en), https://arxiv.org/abs/2609.02046 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9075f6931ffc (Metraipolis archive dc_9075f6931ffc, sha256 35c8cc746155a2eb18aeb3ed041d610238532cb06d7b7d39f0598fbbf5bb4dd6)
BibTeX
@misc{dc9075f6931ffc,
title = {Modeling What Changes: Sparse, Residual World Models for Object-Centric Manipulation — arXiv:2609.02046},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.02046},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9075f6931ffc; sha256 35c8cc746155a2eb18aeb3ed041d610238532cb06d7b7d39f0598fbbf5bb4dd6},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02046. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Monolithic world models predict the entire next state at every step, spending capacity re-predicting the static majority of a scene and injecting error into it. We ask whether explicitly modeling change (a per-object change gate plus a residual delta head that perturbs only the objects the gate flags) is a more effective and interpretable bias for physical prediction and control. On a MuJoCo tabletop pushing benchmark scaling from 3 to 8 objects, the sparse/residual model predicts next-state poses 2.5 to 4.6 times more accurately than a dense multilayer perceptron at 8.6 to 11.1 times fewer parameters, sustains change-detection F1 of 0.80 to 0.87 where the dense baseline is degenerate, transfers across object counts with zero retraining (99.4 percent F1 retention), and reaches about 90 percent of its full-data accuracy with a quarter of the data. In autoregressive rollout it compounds far less error, hugging the no-motion floor while the dense model drifts. Finally, inside a sampling-based planner, prediction-only models fail (though a true-simulator oracle solves the task with the identical planner, confirming the planner is sound), but once featurized and trained for the states a planner visits, the sparse model begins to plan (0.23 plus or minus 0.06 success over three seeds) while the dense monolith stays at zero at every seed. Modeling what changes, rather than re-predicting the whole world, is a simple, effective bias for object-centric physical AI; code, data generators, and all checkpoints will be released upon publication. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.AI). Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.02046 — 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 2026-10-17T03:13:12.498Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).