Metraipolis › archive › dc_f9fdfc81e7e4
Reperesentation Geometry Matters for Planning with JEPA World Models — arXiv:2608.16287
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.16287
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T19:41:34.944Z
- SHA-256 of the archived text
3164a58b918d846fba2ba0fd985b4c816a19e1d74b818afc23c03f59af345f11- Stable URL of this page
https://metraipolis.com/documento/dc_f9fdfc81e7e4
How to cite
Reperesentation Geometry Matters for Planning with JEPA World Models — arXiv:2608.16287 — arXiv (en), https://arxiv.org/abs/2608.16287 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f9fdfc81e7e4 (Metraipolis archive dc_f9fdfc81e7e4, sha256 3164a58b918d846fba2ba0fd985b4c816a19e1d74b818afc23c03f59af345f11)
BibTeX
@misc{dcf9fdfc81e7e4,
title = {Reperesentation Geometry Matters for Planning with JEPA World Models — arXiv:2608.16287},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.16287},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f9fdfc81e7e4; sha256 3164a58b918d846fba2ba0fd985b4c816a19e1d74b818afc23c03f59af345f11},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.16287. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Joint-embedding predictive world models support planning through latent predictions, but unconstrained joint training can collapse distinct observations to identical embeddings. Two prominent strategies for avoiding collapse are to inherit pretrained features, as in DINO-WM, or to learn representations end-to-end with anti-collapse regularization, as in LeWorldModel (LeWM). Yet avoiding collapse does not ensure that latent distances distinguish outcomes in ways that matter for the task. In object manipulation, for example, success depends on the object's position and orientation relative to the goal. Such task-relevant state information can remain accurately decodable while barely influencing latent distance. The resulting planning cost may fail to reflect how close a predicted outcome is to the task goal. In this paper, we propose SCALE (State-CAlibrated Latent Embeddings), a method that correlates sampled pairwise latent distances with distances in task-relevant state space. Added to LeWM's existing objective, SCALE preserves its architecture, requires privileged state only during training, and adds no planning-time computation. We show that SCALE improves planning success over LeWM across manipulation and navigation tasks with multiple solvers and provide a comprehensive analysis of how SCALE reshapes representation geometry to support planning. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2608.16287 — 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 2026-10-18T19:02:44.165Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).