Metraipolis › archive › dc_ea0df6ca4a1d
How Long, Not How Close: A Learned Temporal Metric for Planning in Latent World Models — arXiv:2610.04988
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.04988
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T05:30:50.389Z
- SHA-256 of the archived text
17aca5bcafd8c0fce484b6b1181b9ee7760f3fb96e319dcc98b09cff1d55ba52- Stable URL of this page
https://metraipolis.com/documento/dc_ea0df6ca4a1d
How to cite
How Long, Not How Close: A Learned Temporal Metric for Planning in Latent World Models — arXiv:2610.04988 — arXiv, https://arxiv.org/abs/2610.04988 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ea0df6ca4a1d (Metraipolis archive dc_ea0df6ca4a1d, sha256 17aca5bcafd8c0fce484b6b1181b9ee7760f3fb96e319dcc98b09cff1d55ba52)
BibTeX
@misc{dcea0df6ca4a1d,
title = {How Long, Not How Close: A Learned Temporal Metric for Planning in Latent World Models — arXiv:2610.04988},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.04988},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ea0df6ca4a1d; sha256 17aca5bcafd8c0fce484b6b1181b9ee7760f3fb96e319dcc98b09cff1d55ba52},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04988. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Latent world models plan by rolling a frozen predictor forward under candidate action sequences and ranking the candidates by the latent distance between their imagined end state and the goal. However, this ranking breaks down when the goal lies several plans away, because the latent distance measures how closely an end state resembles the goal rather than how far it remains from reaching it. To address this, we propose TEMPO, a temporal-distance planning objective that leaves the world model untouched, learns only from the demonstrations already used to train it, and adds negligible cost to the planner's search. TEMPO learns a small map of the frozen latent in which the distance between two states of an episode reflects the number of environment steps between them, and blends this distance into the planner's cost. It requires no rewards, policies or success labels and, being a cost rather than a model, applies to frozen world models with one latent vector per state that plan by a latent distance. We evaluate TEMPO on eleven simulated environments (e.g., maze navigation, tabletop pushing, robotic arm control and three-dimensional manipulation) with the LeWM and PLDM planners. With a small MLP that adds at most 0.3% to a plan's arithmetic, TEMPO improves both planners at every goal distance, including the one-plan setting of their evaluations, raises LeWM from 36% to 99% on TwoRoom three plans from the goal, and remains competitive on a broad range of 2D and 3D navigation, reaching and manipulation tasks. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.04988 — 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 2026-10-06T05:27:05.753Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).