Paper Detail
Jianwei Zhang, Sihan Cao, Pengcheng Zheng, Ya Wen, Pei Ke, Kuien Liu, Shen Gao, Wei Dong, Yang Yang, Chaoning Zhang
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 https://github.com/zjw131f1fc/budgeted-online-teaching.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@article{zhang2026when,
title = {When and What to Teach: Budget-Aware Online Adaptation for Web Agents},
author = {Jianwei Zhang and Sihan Cao and Pengcheng Zheng and Ya Wen and Pei Ke and Kuien Liu and Shen Gao and Wei Dong and Yang Yang and Chaoning Zhang},
year = {2026},
abstract = {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 },
url = {https://arxiv.org/abs/2609.05513},
keywords = {cs.AI, cs.CV, cs.LG},
eprint = {2609.05513},
archiveprefix = {arXiv},
}
{}