原文 · 未翻译
Computer Science > Computation and Language
Title:Qwen-AgentWorld: Language World Models for General Agents
Yuxin Zuo
Zikai Xiao
Li Sheng
Fei Huang
Jianhong Tu
Yuxuan Liu
Tianyi Tang
Xiaomeng Hu
Yang Su
Qingfeng Lan
Yantao Liu
Qin Zhu
Yinger Zhang
Bowen Yu
Haiquan Zhao
Haiyang Xu
Jianxin Yang
Jiayang Cheng
Junyang Wang
Lianghao Deng
Mingfeng Xue
Tianyi Bai
Yang Fan
Yubo Ma
Yucheng Li
Zeyu Cui
Zhihai Wang
Zhihui Xie
Zhuorui Ye
An Yang
Dayiheng Liu
Jingren Zhou
Ning Ding
Abstract:A world model predicts environment dynamics based on current observations and actions, serving as a core cognitive mechanism for reasoning and planning. In this work, we investigate how world modeling based on language models can further push the boundaries of general agents. (i) We first focus on building foundation models for agentic environment simulation. We introduce Qwen-AgentWorld-35B-A3B and Qwen-AgentWorld-397B-A17B, the first language world models capable of simulating agentic environments covering 7 domains via long chain-of-thought reasoning. Leveraging more than 10M environment interaction trajectories of 7 domains in real-world environments, we develop Qwen-AgentWorld through a three-stage training pipeline: CPT injects general-purpose world modeling capabilities from the state transition dynamics and augmented professional corpora, SFT activates next-state-prediction reasoning, and RL sharpens simulation fidelity through a tailored framework with hybrid rubric-and-rule rewards. To evaluate language world models, we present AgentWorldBench, a comprehensive benchmark constructed from real-world interactions of 5 frontier models on 9 established benchmarks. Empirical results demonstrate that Qwen-AgentWorld significantly outperforms existing frontier models. (ii) Beyond foundation models, we further investigate two complementary paradigms through which world modeling enhances general agents. First, as a decoupled environment simulator, Qwen-AgentWorld supports scalable and controllable simulation of thousands of real-world environments for agentic RL, yielding gains that surpass real-environment training alone. Second, as a unified agent foundation model, world-model training acts as a highly effective warm-up that improves downstream performance across 7 agentic benchmarks. Code: this https URL
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2606.24597 [cs.CL] |
| (or arXiv:2606.24597v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2606.24597 arXiv-issued DOI via DataCite |
Submission history
From: Fei Huang [
Tue, 23 Jun 2026 13:53:55 UTC (3,883 KB)
Access Paper:
Current browse context:
References & Citations
Bookmark
