Metraipolis › archive › dc_232c41b4367c

Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use From Evolved Harnesses — arXiv:2610.40747

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.40747
License (as archived): CC0-1.0
Archived: 2026-11-02T07:12:32.150Z
SHA-256 of the archived text: 8dc7aa0628eff36b6eab1e57b66cb6a7f797c80b16e4e192090494459252261c
Stable URL of this page: https://metraipolis.com/documento/dc_232c41b4367c

How to cite

Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use From Evolved Harnesses — arXiv:2610.40747 — arXiv (en), https://arxiv.org/abs/2610.40747 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_232c41b4367c (Metraipolis archive dc_232c41b4367c, sha256 8dc7aa0628eff36b6eab1e57b66cb6a7f797c80b16e4e192090494459252261c)

BibTeX

@misc{dc232c41b4367c,
  title = {Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use From Evolved Harnesses — arXiv:2610.40747},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.40747},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_232c41b4367c; sha256 8dc7aa0628eff36b6eab1e57b66cb6a7f797c80b16e4e192090494459252261c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.40747. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Effective computer-use agents (CUAs) should seamlessly combine visual GUI interaction with code and terminal execution within a single policy. We present Mage-VL-Cowork, a 4B hybrid CUA trained from a general multimodally aligned checkpoint through an end-to-end, progressive five-stage curriculum across midtraining, SFT, and RL. To power this curriculum at scale, Mage-CUA-Forge replaces per-task agent construction with reusable, skill-conditioned generation workflows and application-specific preparers, synthesizing 293,806 desktop and web tasks across 6,915 skills while reducing recurring per-task expenditure by 82.6%. Rather than directly distilling a default teacher, we evolve the teacher harness and instantiate specialized CUA harnesses to yield high-quality trajectories alongside stage-specific supervision. Mage-VL-Cowork-4B finally climbs to 66.0% on OSWorld-Verified through different training stages with more efficient reasoning cost, achieving state-of-the-art performance at its parameter size and pushing forward the frontier of on-device intelligence.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.40747 — 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 2026-11-02T07:07:14.179Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).