Metraipolis › archive › dc_9682fa69ccc3

WebPageBench: Event-Level Verification and Controlled UI-Variant Generation for Web Agents — arXiv:2609.35026

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.35026
License (as archived): CC0-1.0
Archived: 2026-10-07T20:58:30.662Z
SHA-256 of the archived text: eb0328a41568b65b9297f3032395118a81e01021fc03345a3c78b83a39c240b1
Stable URL of this page: https://metraipolis.com/documento/dc_9682fa69ccc3

How to cite

WebPageBench: Event-Level Verification and Controlled UI-Variant Generation for Web Agents — arXiv:2609.35026 — arXiv (en), https://arxiv.org/abs/2609.35026 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9682fa69ccc3 (Metraipolis archive dc_9682fa69ccc3, sha256 eb0328a41568b65b9297f3032395118a81e01021fc03345a3c78b83a39c240b1)

BibTeX

@misc{dc9682fa69ccc3,
  title = {WebPageBench: Event-Level Verification and Controlled UI-Variant Generation for Web Agents — arXiv:2609.35026},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.35026},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9682fa69ccc3; sha256 eb0328a41568b65b9297f3032395118a81e01021fc03345a3c78b83a39c240b1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35026. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We present WebPageBench, an open framework for evaluating web agents in which every task is verified from the interface's own event log. Six instrumented mock sites with brand identifiers removed (a marketplace, a bookstore, a grocery service, rail ticketing, hotel search and a document cabinet) emit typed events with parameters as a user or an agent acts. A task declares the events it requires, and success is decided by matching them, with no judge model and no scraping of rendered pages. The same instrumentation supports controlled UI variation: one configuration switch re-renders a task through a different implementation of a single control while the prompt and the success conditions stay completely identical, so sensitivity to interface form can be measured under a fixed task specification. The WebPageBench release consists of three components: 152 tasks, divided into 65 canonical scenarios and 87 control variants across light/dark UI-modes; a common runner evaluated with six browser/DOM harness configurations and five screenshot-only GUI-agent families; and a public leaderboard of 24 model-harness pairs. On the public 152-task leaderboard the gap between what agents declare finished and what the log confirms reaches 41 points (one configuration declares every task finished and satisfies the conditions on 59%).

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.35026 — 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 2026-10-07T20:29:01.756Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).