Metraipolis › archive › dc_fc43608bb33b
Teaching Agents in their Own Style: Blackbox Teacher Distillation for Web Agents — arXiv:2610.19834
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.19834
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T20:02:06.702Z
- SHA-256 of the archived text
b5e511650e75fc20d6ab7a791b639e68ba150fe5dcf84479157e2e74a150d6b2- Stable URL of this page
https://metraipolis.com/documento/dc_fc43608bb33b
How to cite
Teaching Agents in their Own Style: Blackbox Teacher Distillation for Web Agents — arXiv:2610.19834 — arXiv (en), https://arxiv.org/abs/2610.19834 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fc43608bb33b (Metraipolis archive dc_fc43608bb33b, sha256 b5e511650e75fc20d6ab7a791b639e68ba150fe5dcf84479157e2e74a150d6b2)
BibTeX
@misc{dcfc43608bb33b,
title = {Teaching Agents in their Own Style: Blackbox Teacher Distillation for Web Agents — arXiv:2610.19834},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.19834},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fc43608bb33b; sha256 b5e511650e75fc20d6ab7a791b639e68ba150fe5dcf84479157e2e74a150d6b2},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19834. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Frontier web agents using closed models have high inference and computational costs. Smaller open models can be an alternative, but to be performant, they require training supervision from a human or teacher model. Token-level teacher distillation is not an option for closed teachers as it needs access to the model's logits. The standard approach is to imitate teacher demonstrations offline, but it trains the student on states it has never visited, and in a style it does not write in. We introduce WebKD, where the teacher is the editor of the student's failures. Given a failed episode, the teacher finds the first point of mistake, and rewrites the student's reasoning traces, harness states, and actions from that point onward. The editing reveals that the teacher's corrections can be more surprising to the student than the student's own text. WebKD instead aims to edit minimally. In the data space, the teacher writes each correction in the student's style. In the objective space, the reasoning is weighted by its contribution to the action. A role-asymmetric KL anchor allows the corrected action to move freely while keeping the reasoning near the student's prior. On the TimeWarp benchmark, WebKD students gain about 5.5-10.7 pp over behavior cloning, outcome-reward RL, and teacher editing baselines. The performance transfers to unseen versions and to the live web in the OnlineMind2Web benchmark. WebKD opens a new way of distilling black-box models to improve the capabilities of open-weight web agents. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.19834 — 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 2026-10-15T19:35:48.272Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).