Metraipolis › archive › dc_6aedab767443

MAC-v2: Long-horizon model-based value expansion from pixels — arXiv:2611.07245

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.07245
License (as archived)
CC0-1.0
Archived
2026-11-06T14:42:15.162Z
SHA-256 of the archived text
8840da889ac188138269b50c4f3f2bea0e4bf8317a17b4f6f0fb37eab76299a2
Stable URL of this page
https://metraipolis.com/documento/dc_6aedab767443

How to cite

MAC-v2: Long-horizon model-based value expansion from pixels — arXiv:2611.07245 — arXiv (en), https://arxiv.org/abs/2611.07245 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6aedab767443 (Metraipolis archive dc_6aedab767443, sha256 8840da889ac188138269b50c4f3f2bea0e4bf8317a17b4f6f0fb37eab76299a2)
BibTeX
@misc{dc6aedab767443,
  title = {MAC-v2: Long-horizon model-based value expansion from pixels — arXiv:2611.07245},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.07245},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6aedab767443; sha256 8840da889ac188138269b50c4f3f2bea0e4bf8317a17b4f6f0fb37eab76299a2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07245. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We study whether \emph{long-horizon model-based value expansion} can serve as an effective framework in pixel-based offline reinforcement learning (RL). Long-horizon model-based value expansion from pixels is substantially harder than state-based environments, as the agent must learn both the representation and its dynamics, and errors from either can accumulate recursively over imagined trajectories. We introduce MAC-v2, which enables long-horizon model-based expansion from pixels by training a reconstruction-free latent world model with action chunking. By learning representations from prediction and control objectives, MAC-v2 does not need to encode control-irrelevant visual details that can hinder the dynamics prediction. Action chunking further mitigates the prediction error by reducing the number of recursive model predictions required for value expansion. Across diverse visual benchmarks including visual OGBench, MMBench2, Robomimic, and LIBERO-10, MAC-v2 achieves strong performance against model-free and model-based offline RL baselines. Our analyses show that long-horizon model-based value expansion is critical for the performance of MAC-v2, and that both action chunking and latent learning objectives are essential for reliable long-horizon latent imagination.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.07245 — 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 2026-11-06T14:18:42.669Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).