Metraipolis › archive › dc_afc7b2cdf3f4

CRONOS: Benchmarking Non-Episodic Reinforcement Learning for Robotic Manipulation — arXiv:2611.05080

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.05080
License (as archived)
CC0-1.0
Archived
2026-11-04T18:56:02.395Z
SHA-256 of the archived text
0ceb2ea6c40892252bc7b6d8075e04e88412274e659f517ccb1cc155594134b4
Stable URL of this page
https://metraipolis.com/documento/dc_afc7b2cdf3f4

How to cite

CRONOS: Benchmarking Non-Episodic Reinforcement Learning for Robotic Manipulation — arXiv:2611.05080 — arXiv (en), https://arxiv.org/abs/2611.05080 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_afc7b2cdf3f4 (Metraipolis archive dc_afc7b2cdf3f4, sha256 0ceb2ea6c40892252bc7b6d8075e04e88412274e659f517ccb1cc155594134b4)
BibTeX
@misc{dcafc7b2cdf3f4,
  title = {CRONOS: Benchmarking Non-Episodic Reinforcement Learning for Robotic Manipulation — arXiv:2611.05080},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.05080},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_afc7b2cdf3f4; sha256 0ceb2ea6c40892252bc7b6d8075e04e88412274e659f517ccb1cc155594134b4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.05080. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reinforcement learning (RL) is promising for adapting robot policies to unstructured real-world environments. However, standard RL pipelines rely on episodic training with frequent scene resets, which is impractical for real-world deployment due to the need for substantial human intervention. While prior work has studied non-episodic RL in relatively small policies trained from scratch, its behavior when fine-tuning modern large-scale pre-trained vision-language-action (VLA) models remains underexplored. In this work, we introduce CRONOS, a simulation benchmark for RL fine-tuning of VLAs under non-episodic scene resets. CRONOS leverages high-fidelity physics simulators and adopts shared-scene multi-task settings. Using CRONOS, we show that naively fine-tuning pre-trained policies fails in non-episodic settings due to shifts in object configurations and end-effector poses; however, these challenges can be mitigated through targeted reset allocation and by addressing biases in pre-trained models. Moreover, we find that non-episodic training enhances long-horizon manipulation and improves generalization to previously unseen object configurations and task sequences. Project page: this https URL

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Robotics (cs.RO) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.05080 — 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 2026-11-04T18:35:37.545Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).