WebFactory: Automated Compression of Foundational Language Intelligence into Grounded Web Agents

Sicheng Fan, Qingyun Shi, Shengze Xu, Shengbo Cai, Tieyong Zeng, LI LING, Yanyi Shang, Dehan Kong

Keywords: web agent, offline web environment, benchmark, reinforcement learning, synthetic data, GUI grounding

TL;DR: An open-source, fully controllable offline web environment whose built-in site knowledge drives a pipeline to generate executable tasks and high-quality RL data, significantly boosting web-agent performance.

Abstract:

Current paradigms for training GUI agents are fundamentally limited by a reliance on either unsafe, non-reproducible live web interactions or costly, scarce human-crafted data and environments. We argue this focus on data volume overlooks a more critical factor: the efficiency of compressing a large language model's (LLM) latent knowledge into actionable agent behavior. We introduce WebFactory, a novel, fully automated closed-loop reinforcement learning pipeline for GUI agents, systematically compressing LLM-encoded internet intelligence into efficient, grounded actions. Our pipeline features a process of scalable environment synthesis → knowledge-aware task generation → LLM-powered trajectory collection → decomposed reward RL training → systematic agent evaluation.
Remarkably, our agent demonstrates exceptional data efficiency and generalization. Trained on synthetic data from only 10 websites within WebFactory, it achieves performance comparable to GUI agents trained on same amount of human-annotated data from a much larger set of environments. This superior performance is consistent across our internal offline and online transferring benchmarks, where our agent also significantly outperforms the base foundation model.
We further provide critical insights into the "embodiment potential" of different LLM foundations, offering a new axis for model evaluation. This work presents a scalable and cost-effective paradigm for transforming passive internet knowledge into active, grounded intelligence, marking a critical step towards general-purpose interactive agents.

Primary Area: datasets and benchmarks

Submission Number: 13146