Metraipolis › archive › dc_2cfa1253d17d

When and What to Teach: Budget-Aware Online Adaptation for Web Agents — arXiv:2609.05513

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.05513
License (as archived)
CC0-1.0
Archived
2026-10-16T13:58:36.249Z
SHA-256 of the archived text
8ef47e6d31fe04425b5f5b8de4f859e6241fc9dd13c9fc98d944a130a6a5f45e
Stable URL of this page
https://metraipolis.com/documento/dc_2cfa1253d17d

How to cite

When and What to Teach: Budget-Aware Online Adaptation for Web Agents — arXiv:2609.05513 — arXiv (en), https://arxiv.org/abs/2609.05513 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2cfa1253d17d (Metraipolis archive dc_2cfa1253d17d, sha256 8ef47e6d31fe04425b5f5b8de4f859e6241fc9dd13c9fc98d944a130a6a5f45e)
BibTeX
@misc{dc2cfa1253d17d,
  title = {When and What to Teach: Budget-Aware Online Adaptation for Web Agents — arXiv:2609.05513},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.05513},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2cfa1253d17d; sha256 8ef47e6d31fe04425b5f5b8de4f859e6241fc9dd13c9fc98d944a130a6a5f45e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.05513. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Web agents have achieved significant success in automating complex internet tasks but deploying them in real-world environments requires continuous online adaptation. Given that deploying powerful proprietary models remains commercially cost-prohibitive, practitioners must rely on lightweight local models that evolve post-deployment via online teaching from a stronger teacher. However, standard interactive feedback imposes prohibitive costs. We show that conventional trajectory-level preference optimization wastes budget on both unresolvable episodes and redundant execution turns. To resolve these inefficiencies, we propose \textbf{Score-Guided Online Teaching with Budgeted Trajectory Trimming}, a budget-aware framework that systematically orchestrates \textbf{when} and \textbf{what} to teach. Specifically, our framework integrates a solvability-aware teacher gate to dictate \textbf{when} to query the teacher model and a score-guided turn selection mechanism to decide \textbf{what} informative turns to retain. Extensive experiments on MiniWoB and TimeWarp demonstrate that our method achieves comparable first-pass success while reducing teacher calls by 22.6\% and student training compute by 52.1\% on average. Our code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-09 in cs.LG).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.05513 — 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 2026-10-16T13:39:43.110Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).