Metraipolis › archive › dc_77aa4dc99f85

Demonstration-Free Success-Probability Reward Learning for Generalist Robot Policies — arXiv:2609.33653

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.33653
License (as archived)
CC0-1.0
Archived
2026-10-07T21:01:39.426Z
SHA-256 of the archived text
67b600304a93c8cc42ec6a55501a43b3808c1eca968b8392da520bd01c85b56a
Stable URL of this page
https://metraipolis.com/documento/dc_77aa4dc99f85

How to cite

Demonstration-Free Success-Probability Reward Learning for Generalist Robot Policies — arXiv:2609.33653 — arXiv (en), https://arxiv.org/abs/2609.33653 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_77aa4dc99f85 (Metraipolis archive dc_77aa4dc99f85, sha256 67b600304a93c8cc42ec6a55501a43b3808c1eca968b8392da520bd01c85b56a)
BibTeX
@misc{dc77aa4dc99f85,
  title = {Demonstration-Free Success-Probability Reward Learning for Generalist Robot Policies — arXiv:2609.33653},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.33653},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_77aa4dc99f85; sha256 67b600304a93c8cc42ec6a55501a43b3808c1eca968b8392da520bd01c85b56a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.33653. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reinforcement learning (RL) enables generalist robot policies to improve through trial-and-error interaction, yet its effectiveness is fundamentally constrained by sparse task rewards. Existing general-purpose reward models typically alleviate this issue by learning task progress from expert demonstrations, but introduce a distribution mismatch with the mixed-quality rollouts encountered during policy optimization, making their estimates unreliable on suboptimal and failed behaviors from which the policy must learn. In this work, we introduce a demonstration-free reward learning paradigm where dense reward feedback can be learned directly from sparse task outcomes and policy experience. We theoretically show that terminal task outcomes implicitly define dense success-probability feedback at intermediate timesteps, which can be recursively learned through bootstrapping. Based on this insight, we introduce eVTA$_0$, which learns success probabilities from mixed-quality policy rollouts through temporal-difference-style bootstrapping, without expert demonstrations or intermediate annotations. We further introduce RL with Evolving Rewards (RLER), a closed-loop framework that adapts eVTA$_0$ using newly collected rollouts as the policy evolves. Experiments show that eVTA$_0$ provides more informative rewards than state-of-the-art reward models and achieves the best average policy performance across all LIBERO task suites under the same RL training budget, improving success rates by 5.4%-13.8% over the initial policy. In real-world manipulation, RLER further improves overall success rates by 20%-26%, with 35%-36% gains under out-of-distribution conditions. These results demonstrate the effectiveness of demonstration-free reward learning and adapting rewards as the policy evolves. Project webpage: this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Robotics (cs.RO) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.33653 — 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 2026-10-07T20:29:02.718Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).