Metraipolis › archive › dc_3d128ea06ea5

Beyond Action Prediction: Learning Search-Free Planning in Latent Space — arXiv:2610.23615

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.23615
License (as archived): CC0-1.0
Archived: 2026-10-19T08:43:19.045Z
SHA-256 of the archived text: fa3c0625f9ee8087ef1a0314ef1117984525bc9625c8248ee71d3818d14a413b
Stable URL of this page: https://metraipolis.com/documento/dc_3d128ea06ea5

How to cite

Beyond Action Prediction: Learning Search-Free Planning in Latent Space — arXiv:2610.23615 — arXiv (en), https://arxiv.org/abs/2610.23615 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3d128ea06ea5 (Metraipolis archive dc_3d128ea06ea5, sha256 fa3c0625f9ee8087ef1a0314ef1117984525bc9625c8248ee71d3818d14a413b)

BibTeX

@misc{dc3d128ea06ea5,
  title = {Beyond Action Prediction: Learning Search-Free Planning in Latent Space — arXiv:2610.23615},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.23615},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3d128ea06ea5; sha256 fa3c0625f9ee8087ef1a0314ef1117984525bc9625c8248ee71d3818d14a413b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.23615. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Latent world models learn action-conditioned state transitions in a compact representation space, providing a predictive model for visual planning and control. However, planning with these models typically requires costly test-time optimization over action sequences. We propose to replace this online search with a goal-conditioned inverse model that directly predicts actions from current and goal observations. A key challenge is that action prediction alone neither ensures that goal representations encode action-relevant information nor that predicted actions are consistent with the learned latent prediction model. We introduce Hindsight Inverse JEPA Planning (HIJP), a search-free framework that addresses both aspects during training. HIJP relabels future observations along offline trajectories as goals, providing action supervision across multiple temporal horizons. By retaining gradients through the goal representations, this supervision directly shapes the visual representation. HIJP further feeds predicted actions into the latent predictor model and aligns the resulting next-state representations with encoded successor observations, enforcing action-transition consistency. At inference time, the inverse dynamics model and latent predictor models are applied alternately to generate multi-step action sequences without online search. Across four visual control tasks, HIJP achieves 86.8% overall success and a 1.9-16.5x planning speedup over search-based planning.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.23615 — 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 2026-10-19T08:39:10.161Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).