跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 26 天前AI 評分49

WebAnything:從真實網絡構建可擴展的智能體環境

WebAnything: From the Wild Web to Scalable Agent Environments

AI 導讀

WebAnything 可從真實網址全自動重建可重用、以來源為依據的網頁環境,供智能體訓練及評估。系統建立了 244 個網站資產,涵蓋 14 個應用領域及 53 個類別,並支援 12,984 項已執行及獲接納的任務。以 5,856 條互動軌跡微調 Qwen3.5-9B 後,REAL 成功率提升 9.9 個百分點,WebArena-Verified 整體成功率提升 13.33 個百分點。

正文

Published on Oct 27

Authors:

,

,

,

,

,

Abstract

WebAnything automates the reconstruction of reusable, source-grounded web environments from real URLs, enabling scalable agent training and evaluation with verified fidelity.

Web agents require diverse, executable, and resettable website environments for reproducible training and evaluation. Human-assisted mirrors demand per-site engineering, while generative construction can diverge from specified real websites. Capture-and-replay preserves observations, with interaction coverage constrained by captured content and replay capabilities. We present WebAnything, a fully automated framework that reconstructs source-grounded, reusable Web environments from real URLs without per-site manual implementation or repair after initial setup and authorization. An Explorer-Builder collects browser-observable structures, resources, and interaction outcomes; an independent verifier checks frozen candidates and guides local repair or targeted source recapture. We construct 244 website artifacts across 14 application domains and 53 categories, supporting 12,984 executed and accepted tasks. Compared with human-assisted WebHarbor, WebAnything achieves higher aggregate scores on three fidelity metrics, with 12 wins, two ties, and no losses in a blind audit. Fine-tuning Qwen3.5-9B on 5,856 interaction trajectories improves REAL success by 9.9 percentage points, Odysseys-Easy Rubric Avg and Task Avg by 6.25 and 6.32 percentage points, and WebArena-Verified Overall success by 13.33 percentage points under the fixed harness. Code: https://github.com/china-qijizhifeng/Web-Anything

View arXiv page View PDF Project page GitHub 1 Add to collection

Get this paper in your agent:

hf papers read 2610.37786

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.37786 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.37786 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.37786 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co