Metraipolis › archive › dc_0ceedf6cb4b7

NorthCUA Technical Report: Turbocharging Computer-Use Agents through Experience at Scale — arXiv:2611.02102

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.02102
License (as archived)
CC0-1.0
Archived
2026-11-03T18:37:10.912Z
SHA-256 of the archived text
0062384e6048b700b163b3bbc48d54ae76ae1ee37a188037325d00752040a52b
Stable URL of this page
https://metraipolis.com/documento/dc_0ceedf6cb4b7

How to cite

NorthCUA Technical Report: Turbocharging Computer-Use Agents through Experience at Scale — arXiv:2611.02102 — arXiv (en), https://arxiv.org/abs/2611.02102 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0ceedf6cb4b7 (Metraipolis archive dc_0ceedf6cb4b7, sha256 0062384e6048b700b163b3bbc48d54ae76ae1ee37a188037325d00752040a52b)
BibTeX
@misc{dc0ceedf6cb4b7,
  title = {NorthCUA Technical Report: Turbocharging Computer-Use Agents through Experience at Scale — arXiv:2611.02102},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.02102},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0ceedf6cb4b7; sha256 0062384e6048b700b163b3bbc48d54ae76ae1ee37a188037325d00752040a52b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02102. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Scaling computer-use agents from individual actions to complex software work requires experience that connects decisions to complete outcomes. We present NorthCUA, an integrated system for turbocharging computer-use agents through experience at scale. Its Experience Flywheel supplies diverse executable tasks to the NorthCUA Harness, which sustains long interactions, records decisions, and checks task outcomes. The Flywheel uses these records to curate verified trajectories. For unsolved tasks, recursive self-improvement evolves reusable skills and preserves verified milestones so later attempts build on earlier progress. Staged Learning combines this experience with grounding data from the Flywheel to develop GUI competence through mid-training and long-horizon supervised fine-tuning. The policy then learns from fresh attempts in the same Harness through joint desktop-web reinforcement learning, closing the loop between experience production and policy improvement. The resulting Nano, Mini, and Pro family achieves state-of-the-art results among open-source models across GUI grounding, desktop execution, and Web interaction under the reported evaluation settings. For more responsive Web workflows, NorthCUA-Ultrafast pairs direct action selection with open-ended generation on pre-RL Nano, offering a measured quality-latency tradeoff. Together, these advances establish NorthCUA as a high-performing model family and a reusable foundation for agents that carry complex software work from intent to completion.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.02102 — 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 2026-11-03T18:12:54.544Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).