Metraipolis › archive › dc_f6cfbfbfc014

Learning to Plan from Random Exploration — arXiv:2609.38383

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.38383
License (as archived): CC0-1.0
Archived: 2026-10-07T13:45:14.945Z
SHA-256 of the archived text: e84c866e525763181b54291b6814ba6a3d6dcd13d0fc491cece368dab68bd6b1
Stable URL of this page: https://metraipolis.com/documento/dc_f6cfbfbfc014

How to cite

Learning to Plan from Random Exploration — arXiv:2609.38383 — arXiv (en), https://arxiv.org/abs/2609.38383 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f6cfbfbfc014 (Metraipolis archive dc_f6cfbfbfc014, sha256 e84c866e525763181b54291b6814ba6a3d6dcd13d0fc491cece368dab68bd6b1)

BibTeX

@misc{dcf6cfbfbfc014,
  title = {Learning to Plan from Random Exploration — arXiv:2609.38383},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.38383},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f6cfbfbfc014; sha256 e84c866e525763181b54291b6814ba6a3d6dcd13d0fc491cece368dab68bd6b1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38383. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Random exploration reveals how an environment can be traversed before a goal is specified. Can this experience support long-range planning without policy-improvement training? Our random-walk analysis explains what temporal relations contain: short horizons reveal geodesic geometry in the diffusion limit, while longer horizons reveal connectivity between regions before mixing removes these distinctions. We learn these relations with a conditional energy-based model that estimates temporal log-density ratios through horizon-conditioned embeddings. The model is trained on observation pairs by noise-contrastive estimation, without action or reward labels. The planner queries these learned relations at different horizons as it moves toward the goal. At test time, a separate local dynamics model predicts candidate action outcomes, and the temporal model evaluates their progress toward the goal by selecting or aggregating estimated improvements across horizons. The agent executes one action and replans with both models fixed. Experiments demonstrate long-range maze planning from random exploration using states and images. Learned score fields, embedding probes, and planned routes exhibit properties of a multiscale cognitive map. We further demonstrate egocentric navigation from random exploration and manipulation planning from suboptimal data.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Robotics (cs.RO); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2609.38383 — 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 2026-10-07T13:18:10.196Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).