Metraipolis › archive › dc_aa7d8d7018ea
PAVE: Predictive Alignment and Value-Guided Evolution for World-Action Policies — arXiv:2608.30378
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.30378
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T07:52:37.370Z
- SHA-256 of the archived text
08fe860fa9e12fe2049774c4bc507aee8cd4cf72d01c067879be182fafdc9196- Stable URL of this page
https://metraipolis.com/documento/dc_aa7d8d7018ea
How to cite
PAVE: Predictive Alignment and Value-Guided Evolution for World-Action Policies — arXiv:2608.30378 — arXiv (en), https://arxiv.org/abs/2608.30378 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_aa7d8d7018ea (Metraipolis archive dc_aa7d8d7018ea, sha256 08fe860fa9e12fe2049774c4bc507aee8cd4cf72d01c067879be182fafdc9196)
BibTeX
@misc{dcaa7d8d7018ea,
title = {PAVE: Predictive Alignment and Value-Guided Evolution for World-Action Policies — arXiv:2608.30378},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.30378},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_aa7d8d7018ea; sha256 08fe860fa9e12fe2049774c4bc507aee8cd4cf72d01c067879be182fafdc9196},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.30378. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Direct vision-language-action policies generate continuous robot actions efficiently, but standard behavior cloning leaves two complementary gaps: their representations are not explicitly required to describe how the scene evolves over multiple time scales, and deployment trajectories of unequal quality are often reused without separating useful dynamics from undesirable behavior. We introduce \method, a direct world-action policy that combines outcome-agnostic predictive learning with outcome-aware policy improvement. \method first retains a local fixed-offset JEPA objective and adds trajectory-relative multi-horizon transition alignment at 25%, 50%, 75%, and 100% of the remaining episode. These training-only targets require the current policy representation to preserve both local physical changes and longer-range task progress, without supplying explicit future tokens to the action head. \method then trains an independent distributional value critic on cumulative deployment trajectories, computes action-chunk-aligned $N$-step advantages, and converts them into positive, negative, or null text conditions for a flow-matching actor. Thus, every valid trajectory can teach what physically happened, while the actor is deployed only under the condition associated with relatively better actions. The multi-horizon predictor and critic are removed from online execution, preserving direct action generation from the current observation, language instruction, and proprioception. \redclaim{Across the three simulation benchmarks, \method achieves the strongest overall performance while preserving the direct actor's online execution path.}
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI).
Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.30378 — 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 2026-10-17T07:28:26.859Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).