Metraipolis › archive › dc_600d5ecec67f

Trusting Imagined Futures: Auditing Imagined Rewards for VLA Post-Training — arXiv:2611.04873

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.04873
License (as archived): CC0-1.0
Archived: 2026-11-04T12:57:36.443Z
SHA-256 of the archived text: ee78a075c46fdc155a7c4055cb1298fdfc6bc1e63cf46722c461903fa7f5721f
Stable URL of this page: https://metraipolis.com/documento/dc_600d5ecec67f

How to cite

Trusting Imagined Futures: Auditing Imagined Rewards for VLA Post-Training — arXiv:2611.04873 — arXiv (en), https://arxiv.org/abs/2611.04873 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_600d5ecec67f (Metraipolis archive dc_600d5ecec67f, sha256 ee78a075c46fdc155a7c4055cb1298fdfc6bc1e63cf46722c461903fa7f5721f)

BibTeX

@misc{dc600d5ecec67f,
  title = {Trusting Imagined Futures: Auditing Imagined Rewards for VLA Post-Training — arXiv:2611.04873},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04873},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_600d5ecec67f; sha256 ee78a075c46fdc155a7c4055cb1298fdfc6bc1e63cf46722c461903fa7f5721f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04873. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Video world models enable candidate actions to be evaluated through imagined outcomes, but high predicted rewards can favor trajectories with poor execution quality. We introduce Trusting Imagined Futures (TIF), an offline supervision selection method for vision-language-action (VLA) post-training that audits predicted rewards using execution feedback. A multidimensional reward model predicts a utility score for candidate ranking and five dimensions of execution quality. An auditor uses reward predictions and uncertainty to identify overestimation patterns learned from paired imagined videos and real executions. TIF accepts the candidate with the highest predicted utility when its advantage over the existing trajectory exceeds a margin and its audit risk passes a threshold; otherwise, it retains the existing supervision. The retained real trajectories then supervise behavior cloning. Across four LIBERO suites and three training seeds, TIF achieves a mean success rate of 76.25%, improving over training on the existing trajectories by 3.00 percentage points with the same trajectory count and training budget. Adding audit gating yields 1.47 percentage points with the reward model and advantage margin held fixed. The gains persist against advantage ranking and logistic regression risk selection at matched replacement counts. On recorded videos from three real robot tasks, TIF increases clean completion among selected trajectories from 58.33% to 70.00% across 60 candidate pools.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.04873 — 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 2026-11-04T12:33:47.253Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).