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, and 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 the same amount of human-annotated data from a much larger set of environments. This superior performance is consistent across our internal offline and online transfer 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.
Evaluating Cognitive Age Alignment in Interactive AI Agents
University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);Shenzhen Children's Hospital(深圳儿童医院);Peking University(北京大学);Hong Kong Polytechnic University(香港理工大学)
Transforming the Voice of the Customer: Large Language Models for Identifying Customer Needs
Kellogg School of Management, Northwestern University(西北大学凯洛格管理学院);MIT Management School, Massachusetts Institute of Technology(麻省理工学院斯隆管理学院);Applied Marketing Science, Inc.(应用营销科学公司)
Guided Prompt Evolution for Vision-Language Models Adaptation
Tsinghua Shenzhen International Graduate School, Tsinghua University(清华大学深圳国际研究生院,清华大学);Sun Yat-sen University(中山大学);Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳))
NaiAD: Initiate Data-Driven Research for LLM Advertising
Tsinghua University(清华大学);College of AI(人工智能学院);Department of Literature, Arts and Communication(文学、艺术与传播系);Anhui International Studies University(安徽国际关系大学);State Key Laboratory of General Artificial Intelligence, BIGAI(通用人工智能国家重点实验室,BIGAI)
Bidirectional Curriculum Generation: A Multi-Agent Framework for Data-Efficient Mathematical Reasoning
Zhejiang University(浙江大学);University of Macau(澳门大学);Peking University(北京大学);Harbin Institute of Technology (Shenzhen)(哈尔滨工业大学(深圳));Wuhan University(武汉大学);Shanghai Artificial Intelligence Laboratory(上海人工智能实验室)