Metraipolis › archive › dc_c3cdb70fb3fd
Learn How to Act from Your Own Interactions: On-Policy Self-Distillation for GUI Agents — arXiv:2609.27307
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.27307
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T09:11:18.814Z
- SHA-256 of the archived text
47334359b76ccbee00d91f2225a39b662f2cfa3cb826f58df0d1e62d82d3605a- Stable URL of this page
https://metraipolis.com/documento/dc_c3cdb70fb3fd
How to cite
Learn How to Act from Your Own Interactions: On-Policy Self-Distillation for GUI Agents — arXiv:2609.27307 — arXiv (en), https://arxiv.org/abs/2609.27307 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c3cdb70fb3fd (Metraipolis archive dc_c3cdb70fb3fd, sha256 47334359b76ccbee00d91f2225a39b662f2cfa3cb826f58df0d1e62d82d3605a)
BibTeX
@misc{dcc3cdb70fb3fd,
title = {Learn How to Act from Your Own Interactions: On-Policy Self-Distillation for GUI Agents — arXiv:2609.27307},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.27307},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c3cdb70fb3fd; sha256 47334359b76ccbee00d91f2225a39b662f2cfa3cb826f58df0d1e62d82d3605a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27307. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Graphical User Interface (GUI) agents enable the fulfillment of complex user instructions through multi-turn interactions with software environments, requiring step-wise reasoning and long-horizon memory to guide actions and retain task-relevant information, respectively. Recent on-policy self-distillation (OPSD) methods have achieved strong performance on GUI grounding, a foundational subtask for GUI agents, owing to dense token-level supervision from privilege-conditioned self-teachers. However, extending existing OPSD methods to multi-turn GUI agents is hindered by self-teachers' limited privilege-following ability and insufficient privileged guidance. In this paper, we introduce GUI-SD-v2, the next version of GUI-SD, which extends OPSD from GUI grounding to multi-turn GUI interaction and addresses key limitations through a two-stage training framework. Specifically, GUI-SD-v2 first strengthens privilege following by jointly optimizing rollouts with and without privileged guidance from the same GUI states. Furthermore, it selectively distills step-specific reasoning and memory guidance through a privilege-conditioned self-teacher, supporting action decisions and the retention of task-relevant information for subsequent interactions. Extensive experiments on two representative GUI agent benchmarks, AndroidWorld and MobileWorld, show that GUI-SD-v2 compares favorably with existing OPSD baselines while consistently outperforming the evaluated state-of-the-art methods in both Pass@1 and Pass@3 success rates. Code and training data will be publicly released. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.27307 — 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 2026-10-08T08:13:26.498Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).