Metraipolis › archive › dc_4690f4cd28f3
Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents — arXiv:2609.17653
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.17653
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T15:04:22.176Z
- SHA-256 of the archived text
9cc8905c5585af936df839f50153a7caa4389510b8d8511b36e96e7ca2a4f88a- Stable URL of this page
https://metraipolis.com/documento/dc_4690f4cd28f3
How to cite
Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents — arXiv:2609.17653 — arXiv (en), https://arxiv.org/abs/2609.17653 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4690f4cd28f3 (Metraipolis archive dc_4690f4cd28f3, sha256 9cc8905c5585af936df839f50153a7caa4389510b8d8511b36e96e7ca2a4f88a)
BibTeX
@misc{dc4690f4cd28f3,
title = {Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents — arXiv:2609.17653},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.17653},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4690f4cd28f3; sha256 9cc8905c5585af936df839f50153a7caa4389510b8d8511b36e96e7ca2a4f88a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.17653. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
GUI agents execute long-horizon tasks on dynamic graphical user interfaces, where pop-ups, delayed loads, and relocated widgets routinely invalidate plans fixed before execution. Recent agent-skill frameworks encapsulate reusable procedural knowledge to mitigate this, yet existing skill designs are largely developed without targeting GUI execution dynamics and treat skills as static artifacts produced before deployment rather than living procedural knowledge that improves through it. We argue that what GUI agents need is not better static skills, but skills that can be revised from execution feedback at deployment time, without additional training. We propose \textbf{EvoSkill-GUI}, a training-free framework in which each skill is a structured multi-file package containing retrieval metadata, executable plans, backup localization, failure-recovery rules, accessibility utilities, and failure cases. EvoSkill-GUI operates through a \textbf{\emph{reflect-revise-reuse}} loop: the executor performs instant in-rollout revisions, an isolated critic diagnoses failed trajectories under strict information isolation, and the executor edits specific skill files through a restricted tool interface. Across MobileWorld, AndroidWorld, and OSWorld, three mainstream GUI benchmarks spanning mobile and desktop platforms, EvoSkill-GUI consistently improves multiple base models without any training, with maximum gains of $+16.2\%$, $+6.0\%$, and $+10.5\%$ respectively, and evolved skill libraries continue to benefit related tasks rather than being rebuilt from scratch. Our code is available at this https URL .
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-17 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.17653 — 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 2026-10-15T14:08:19.053Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-17?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).