Metraipolis › archive › dc_bb096d0fcd33

CUA-Sandbox: Efficient Environments for Computer-Use Agent Reinforcement Learning — arXiv:2609.32750

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.32750
License (as archived): CC0-1.0
Archived: 2026-10-07T20:56:01.094Z
SHA-256 of the archived text: a402a2538970009f7b605b4b36e97a4b32177cf380c80977fbc0eb970975f204
Stable URL of this page: https://metraipolis.com/documento/dc_bb096d0fcd33

How to cite

CUA-Sandbox: Efficient Environments for Computer-Use Agent Reinforcement Learning — arXiv:2609.32750 — arXiv (en), https://arxiv.org/abs/2609.32750 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bb096d0fcd33 (Metraipolis archive dc_bb096d0fcd33, sha256 a402a2538970009f7b605b4b36e97a4b32177cf380c80977fbc0eb970975f204)

BibTeX

@misc{dcbb096d0fcd33,
  title = {CUA-Sandbox: Efficient Environments for Computer-Use Agent Reinforcement Learning — arXiv:2609.32750},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.32750},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bb096d0fcd33; sha256 a402a2538970009f7b605b4b36e97a4b32177cf380c80977fbc0eb970975f204},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.32750. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reinforcement learning enables computer-use agents to improve through interaction with real software environments, including websites and desktop applications. However, conventional deployments replicate an initialized runtime for each independent rollout, even when trajectories use the same software, incurring repeated memory and initialization costs as the number of parallel environments grows. Does an independent computer-use environment require an independent execution runtime? Our key observation is that trajectories require independent mutable state, while initialized application runtimes can be reused across concurrently evolving environments, making state the natural unit of environment independence. Guided by this observation, we introduce CUA-Sandbox, which separates private state capsules from shared runtimes through state-scoped execution and transactional lifecycle operations, including resets and branches, while retaining the original software interfaces and task evaluators. Experiments show comparable or improved task success relative to Docker, while substantially reducing rollout and resource costs. CUA-Sandbox achieves up to a 6.20x increase in rollout throughput, a 9.2x reduction in per-environment memory, and a 504x reduction in incremental storage.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.32750 — 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 2026-10-07T20:29:00.953Z 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).