Metraipolis › archive › dc_85dd24ffe1a0
The Objective Is the Bottleneck: Latent World Models Encode What Their Planners Cannot Use — arXiv:2608.12959
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.12959
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T04:18:16.910Z
- SHA-256 of the archived text
9b75f43f5b96429c7f384c32b809602c6a4bb8b6e7dc33a747cf83db418899bb- Stable URL of this page
https://metraipolis.com/documento/dc_85dd24ffe1a0
How to cite
The Objective Is the Bottleneck: Latent World Models Encode What Their Planners Cannot Use — arXiv:2608.12959 — arXiv (en), https://arxiv.org/abs/2608.12959 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_85dd24ffe1a0 (Metraipolis archive dc_85dd24ffe1a0, sha256 9b75f43f5b96429c7f384c32b809602c6a4bb8b6e7dc33a747cf83db418899bb)
BibTeX
@misc{dc85dd24ffe1a0,
title = {The Objective Is the Bottleneck: Latent World Models Encode What Their Planners Cannot Use — arXiv:2608.12959},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.12959},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_85dd24ffe1a0; sha256 9b75f43f5b96429c7f384c32b809602c6a4bb8b6e7dc33a747cf83db418899bb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12959. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Latent world models are judged by how well they predict, so when planning fails at long horizons the natural reading is that the predictor degrades. On a reproduction of LeWorldModel on TwoRoom we show the binding constraint is the planner's objective instead. The predictor is not the limit: its imagined state seventy-five environment steps ahead is still only 0.189 as wrong as assuming the world froze, while the planner never imagines beyond twenty-five. The objective is. Cross-entropy-method planning minimises squared latent distance, which tracks true distance at r = 0.426, saturates by about eighty arena units and decreases beyond a hundred and twenty, so moving away from the goal can lower the cost. The information is present throughout: a ridge probe recovers position from the frozen embedding at R^2 0.9922. The pathology is the method's, not one reimplementation's. It is present in the authors' released weights, and across four checkpoints long-horizon success rank-orders exactly with metric quality and inversely with prediction accuracy. Replacing only the objective, with nothing retrained and no GPU, lifts goals reached at offset 100 from 26.0% to 98.0%, equals the 98.0% at offset 25, and reaches 92.0% under a third of the budget: planning stops depending on the horizon. The best cost is not the most accurate. A head learned from frame separation alone predicts spatial distance worse than a position probe (r = 0.819 against 0.9897) yet plans better, charging 24% more to cross the environment's dividing wall where squared latent distance charges 4% less. It has learned reachability, not proximity. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.12959 — 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 2026-10-19T04:13:06.725Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).