---
title: "Qwen-AgentWorld: Language World Models for General Agents"
description: "The Qwen Team builds such a model, Qwen-AgentWorld. 123 references. Among them: 0."
authors: ["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"]
year: 2026
venue: "arXiv:2606.24597"
source: "https://arxiv.org/abs/2606.24597"
doi: "10.48550/arxiv.2606.24597"
arxiv: "2606.24597"
licence: "arXiv — non-exclusive licence to distribute"
licence_url: "http://arxiv.org/licenses/nonexclusive-distrib/1.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/zuo-2026-qwen-agentworld-language-world/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zuo-2026-qwen-agentworld-language-world/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/40/) / Qwen-AgentWorld: Language World Models for General Agents

# Qwen-AgentWorld: Language World Models for General Agents

**Authors:** [Yuxin Zuo](https://md.research.elamur.ai/free/en/authors/yuxin-zuo/), [Zikai Xiao](https://md.research.elamur.ai/free/en/authors/zikai-xiao/), [Li Sheng](https://md.research.elamur.ai/free/en/authors/li-sheng/), [Fei Huang](https://md.research.elamur.ai/free/en/authors/fei-huang/), [Jianhong Tu](https://md.research.elamur.ai/free/en/authors/jianhong-tu/), [Yuxuan Liu](https://md.research.elamur.ai/free/en/authors/yuxuan-liu/), [Tianyi Tang](https://md.research.elamur.ai/free/en/authors/tianyi-tang/), [Xiaomeng Hu](https://md.research.elamur.ai/free/en/authors/xiaomeng-hu/), [Yang Su](https://md.research.elamur.ai/free/en/authors/yang-su/), Qingfeng Lan, [Yantao Liu](https://md.research.elamur.ai/free/en/authors/yantao-liu/), Qin Zhu, Yinger Zhang, [Bowen Yu](https://md.research.elamur.ai/free/en/authors/bowen-yu/), Haiquan Zhao, [Haiyang Xu](https://md.research.elamur.ai/free/en/authors/haiyang-xu/), Jianxin Yang, Jiayang Cheng, Junyang Wang, [Lianghao Deng](https://md.research.elamur.ai/free/en/authors/lianghao-deng/), [Mingfeng Xue](https://md.research.elamur.ai/free/en/authors/mingfeng-xue/), Tianyi Bai, [Yang Fan](https://md.research.elamur.ai/free/en/authors/yang-fan/), [Yubo Ma](https://md.research.elamur.ai/free/en/authors/yubo-ma/), [Yucheng Li](https://md.research.elamur.ai/free/en/authors/yucheng-li/), [Zeyu Cui](https://md.research.elamur.ai/free/en/authors/zeyu-cui/), Zhihai Wang, Zhihui Xie, Zhuorui Ye, [An Yang](https://md.research.elamur.ai/free/en/authors/an-yang/), [Dayiheng Liu](https://md.research.elamur.ai/free/en/authors/dayiheng-liu/), [Jingren Zhou](https://md.research.elamur.ai/free/en/authors/jingren-zhou/) and [Ning Ding](https://md.research.elamur.ai/free/en/authors/ning-ding/) · **Year:** 2026 · **Venue:** arXiv:2606.24597 · **Licence:** [arXiv — non-exclusive licence to distribute](http://arxiv.org/licenses/nonexclusive-distrib/1.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2606.24597) · [PDF](https://arxiv.org/pdf/2606.24597) · [DOI 10.48550/arxiv.2606.24597](https://doi.org/10.48550/arxiv.2606.24597) · [arXiv 2606.24597](https://arxiv.org/abs/2606.24597) · [OpenAlex](https://openalex.org/W7165749951)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
LLM agent research has mostly focused on the policy, which picks actions from states. The complementary piece is a world model that predicts the next state from a state and an action, and agents working in text-based environments still lack a general one. The Qwen Team builds such a model, Qwen-AgentWorld. It forecasts environment responses using long chain-of-thought reasoning across seven agentic domains. Training draws on over 10M real interaction trajectories and runs in three stages:

- **CPT** instills transition dynamics and domain knowledge.
- **SFT** elicits step-by-step next-state reasoning.
- **RL** with mixed rubric and rule rewards improves simulation accuracy.

To measure simulation quality, the authors release AgentWorldBench, built from real agent runs on established benchmarks. On it, Qwen-AgentWorld beats existing frontier models. The paper then tests two uses of the model:

1. **Separate simulator for agentic RL.** Controllable perturbations allow training that exceeds training in real environments alone.
2. **Initialization for the agent itself.** World-model training serves as a warm-up that raises downstream agent scores.

This suggests that predicting the environment is a further way to scale general agents, alongside real-environment interaction.

## Key points
- Two model sizes, Qwen-AgentWorld-35B-A3B and Qwen-AgentWorld-397B-A17B, cover seven domains: MCP, Search, Terminal, SWE, Android, Web and OS. In the GUI domains, states are text accessibility trees and view hierarchies, not pixels.
- AgentWorldBench uses trajectories from 5 frontier models (e.g., Claude Opus 4.6) on 9 benchmarks, including Tool Decathlon, Terminal-Bench 1.0 & 2.0 and OSWorld-Verified. Outputs are judged against ground truth on 5 rubric dimensions, with additional rule-based verifiers.
- As a simulator, the model emulates about 4k OpenClaw environments, improving results on Claw-Eval and QwenClawBench. Controlled perturbations, such as returning partial results, boost Tool Decathlon, MCPMark and WideSearch beyond real-environment RL.
- Used as a warm-up, world-model training helps on 7 benchmarks: Terminal-Bench 2.0, SWE-Bench Verified, SWE-Bench Pro, BFCL v4, Claw-Eval, QwenClawBench and WideSearch.
- Planned future directions are agent–world-model co-evolution, multimodal GUI modeling, adaptive routing between simulated and real environments, and on-the-fly tool synthesis.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

📚 Shelves (2): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works

## Cite

Cite as: 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 and Ning Ding. 2026. Qwen-AgentWorld: Language World Models for General Agents. arXiv:2606.24597.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{zuo2026qwen,
  title = {{Qwen-AgentWorld: Language World Models for General Agents}},
  author = {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 and Ning Ding},
  year = {2026},
  doi = {10.48550/arxiv.2606.24597},
  eprint = {2606.24597},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2606.24597}
}
```

- [← previous QurSci-Onto: A Hierarchical Ontology and Dataset for Scientific Exegesis in the Quran](https://md.research.elamur.ai/free/en/papers/rashid-2026-qursci-onto-hierarchical-ontology-dataset/)
- [next → Qwen3.8-Omni: Towards Native Omni-Modal Agents](https://md.research.elamur.ai/free/en/papers/qwen-team-2026-qwen3-8-omni-native/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/40/)

## References

123 references, 13 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Arslan Ali, Junjie Bai, Maciej Bala, Yogesh Balaji, Aaron Blakeman, Tiffany Cai, Jiaxin Cao, Tianshi Cao, Elizabeth Cha, Yu-Wei Chao, et al. World simulation with video foundation models for physical ai. arXiv preprint arXiv:2511.00062, 2025. [find](https://scholar.google.com/scholar?q=Arslan+Ali%2C+Junjie+Bai%2C+Maciej+Bala%2C+Yogesh+Balaji%2C+Aaron+Blakeman%2C+Tiffany+Cai%2C+Jiaxin+Cao%2C+Tianshi+Cao%2C+Elizabeth+Cha%2C+Yu%2DWei+Chao%2C+et+al%2E+World+simulation+with+video+foundation+models+for+physical+)
2. Anthropic. System card: Claude sonnet 4.5, 2025. URL [https://assets.anthropic.com/m/12f214efcc2](https://assets.anthropic.com/m/12f214efcc2) f457a/original/Claude-Sonnet-4-5-System-Card.pdf. [↗](https://assets.anthropic.com/m/12f214efcc2)
3. Anthropic. System card: Claude opus 4.6, 2026a. URL [https://www-cdn.anthropic.com/14e4fb01875d2](https://www-cdn.anthropic.com/14e4fb01875d2) a69f646fa5e574dea2b1c0ff7b5.pdf. [↗](https://www-cdn.anthropic.com/14e4fb01875d2)
4. Anthropic. System card: Claude opus 4.8, 2026b. URL [https://www-cdn.anthropic.com/0f0c97ad20d80](https://www-cdn.anthropic.com/0f0c97ad20d80) 05706296bd92aa1c27c6b2f4f61.pdf. [↗](https://www-cdn.anthropic.com/0f0c97ad20d80)
5. Anthropic. System card: Claude sonnet 4.6, 2026c. URL [https://www-cdn.anthropic.com/bbd8ef16d70](https://www-cdn.anthropic.com/bbd8ef16d70) b7a1665f14f306ee88b53f686aa75.pdf. [↗](https://www-cdn.anthropic.com/bbd8ef16d70)
6. Mido Assran, Adrien Bardes, David Fan, Quentin Garrido, Russell Howes, Matthew Muckley, Ammar Rizvi, Claire Roberts, Koustuv Sinha, Artem Zholus, et al. V-jepa 2: Self-supervised video models enable understanding, prediction and planning. arXiv preprint arXiv:2506.09985, 2025. — [arXiv:2506.09985](https://arxiv.org/abs/2506.09985) [find](https://scholar.google.com/scholar?q=Mido+Assran%2C+Adrien+Bardes%2C+David+Fan%2C+Quentin+Garrido%2C+Russell+Howes%2C+Matthew+Muckley%2C+Ammar+Rizvi%2C+Claire+Roberts%2C+Koustuv+Sinha%2C+Artem+Zholus%2C+et+al%2E+V%2Djepa+2%3A+Self%2Dsupervised+video+models+enable+u)
7. Hao Bai, Alexey Taymanov, Tong Zhang, Aviral Kumar, and Spencer Whitehead. Webgym: Scaling training environments for visual web agents with realistic tasks. arXiv preprint arXiv:2601.02439, 2026. — [arXiv:2601.02439](https://arxiv.org/abs/2601.02439) [find](https://scholar.google.com/scholar?q=Hao+Bai%2C+Alexey+Taymanov%2C+Tong+Zhang%2C+Aviral+Kumar%2C+and+Spencer+Whitehead%2E+Webgym%3A+Scaling+training+environments+for+visual+web+agents+with+realistic+tasks%2E+arXiv+preprint+arXiv%3A2601%2E02439%2C+2026%2E+%E2%80%94+%5Ba)
8. Philip J. Ball, Jakob Bauer, Frank Belletti, Bethanie Brownfield, Ariel Ephrat, Shlomi Fruchter, Agrim Gupta, Kristian Holsheimer, Aleksander Holynski, Jiri Hron, Christos Kaplanis, Marjorie Limont, Matt McGill, Yanko Oliveira, Jack Parker-Holder, Frank Perbet, Guy Scully, Jeremy Shar, Stephen Spenc… [find](https://scholar.google.com/scholar?q=Philip+J%2E+Ball%2C+Jakob+Bauer%2C+Frank+Belletti%2C+Bethanie+Brownfield%2C+Ariel+Ephrat%2C+Shlomi+Fruchter%2C+Agrim+Gupta%2C+Kristian+Holsheimer%2C+Aleksander+Holynski%2C+Jiri+Hron%2C+Christos+Kaplanis%2C+Marjorie+Limont%2C+M)
9. Shihao Cai, Runnan Fang, Jialong Wu, Baixuan Li, Xinyu Wang, Yong Jiang, Liangcai Su, Liwen Zhang, Wenbiao Yin, Zhen Zhang, et al. Autoforge: Automated environment synthesis for agentic reinforcement learning. arXiv preprint arXiv:2512.22857, 2025. — [arXiv:2512.22857](https://arxiv.org/abs/2512.22857) [find](https://scholar.google.com/scholar?q=Shihao+Cai%2C+Runnan+Fang%2C+Jialong+Wu%2C+Baixuan+Li%2C+Xinyu+Wang%2C+Yong+Jiang%2C+Liangcai+Su%2C+Liwen+Zhang%2C+Wenbiao+Yin%2C+Zhen+Zhang%2C+et+al%2E+Autoforge%3A+Automated+environment+synthesis+for+agentic+reinforcement+)
10. Yilin Cao, Yufeng Zhong, Zhixiong Zeng, Liming Zheng, Jing Huang, Haibo Qiu, Peng Shi, Wenji Mao, and Wan Guanglu. Mobiledreamer: Generative sketch world model for gui agent. arXiv preprint arXiv:2601.04035, 2026a. — [arXiv:2601.04035](https://arxiv.org/abs/2601.04035) [find](https://scholar.google.com/scholar?q=Yilin+Cao%2C+Yufeng+Zhong%2C+Zhixiong+Zeng%2C+Liming+Zheng%2C+Jing+Huang%2C+Haibo+Qiu%2C+Peng+Shi%2C+Wenji+Mao%2C+and+Wan+Guanglu%2E+Mobiledreamer%3A+Generative+sketch+world+model+for+gui+agent%2E+arXiv+preprint+arXiv%3A2601)
11. Yuan Cao, Dezhi Ran, Mengzhou Wu, Yuzhe Guo, Xin Chen, Ang Li, Gang Cao, Gong Zhi, Hao Yu, Linyi Li, et al. Gui-genesis: Automated synthesis of efficient environments with verifiable rewards for gui agent post-training. arXiv preprint arXiv:2602.14093, 2026b. — [arXiv:2602.14093](https://arxiv.org/abs/2602.14093) [find](https://scholar.google.com/scholar?q=Yuan+Cao%2C+Dezhi+Ran%2C+Mengzhou+Wu%2C+Yuzhe+Guo%2C+Xin+Chen%2C+Ang+Li%2C+Gang+Cao%2C+Gong+Zhi%2C+Hao+Yu%2C+Linyi+Li%2C+et+al%2E+Gui%2Dgenesis%3A+Automated+synthesis+of+efficient+environments+with+verifiable+rewards+for+gui+a)
12. Hyungjoo Chae, Namyoung Kim, Kai Ong, Minju Gwak, Gwanwoo Song, Jihoon Kim, Sunghwan Kim, Dongha Lee, and Jinyoung Yeo. Web agents with world models: Learning and leveraging environment dynamics in web navigation. In International Conference on Learning Representations, volume 2025, pp. 63707–63738,… [find](https://scholar.google.com/scholar?q=Hyungjoo+Chae%2C+Namyoung+Kim%2C+Kai+Ong%2C+Minju+Gwak%2C+Gwanwoo+Song%2C+Jihoon+Kim%2C+Sunghwan+Kim%2C+Dongha+Lee%2C+and+Jinyoung+Yeo%2E+Web+agents+with+world+models%3A+Learning+and+leveraging+environment+dynamics+in+we)
13. Hyungjoo Chae, Jungsoo Park, and Alan Ritter. Safe and scalable web agent learning via recreated websites. arXiv preprint arXiv:2603.10505, 2026. — [arXiv:2603.10505](https://arxiv.org/abs/2603.10505) [find](https://scholar.google.com/scholar?q=Hyungjoo+Chae%2C+Jungsoo+Park%2C+and+Alan+Ritter%2E+Safe+and+scalable+web+agent+learning+via+recreated+websites%2E+arXiv+preprint+arXiv%3A2603%2E10505%2C+2026%2E+%E2%80%94+%5BarXiv%3A2603%2E10505%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2603%2E10505%29)
14. Mouxiang Chen, Lei Zhang, Yunlong Feng, Xuwu Wang, Wenting Zhao, Ruisheng Cao, Jiaxi Yang, Jiawei Chen, Mingze Li, Zeyao Ma, et al. Swe-universe: Scale real-world verifiable environments to millions. arXiv preprint arXiv:2602.02361, 2026. — [arXiv:2602.02361](https://arxiv.org/abs/2602.02361) [find](https://scholar.google.com/scholar?q=Mouxiang+Chen%2C+Lei+Zhang%2C+Yunlong+Feng%2C+Xuwu+Wang%2C+Wenting+Zhao%2C+Ruisheng+Cao%2C+Jiaxi+Yang%2C+Jiawei+Chen%2C+Mingze+Li%2C+Zeyao+Ma%2C+et+al%2E+Swe%2Duniverse%3A+Scale+real%2Dworld+verifiable+environments+to+millions%2E+)
15. Shiqi Chen, Tongyao Zhu, Zian Wang, Jinghan Zhang, Kangrui Wang, Siyang Gao, Teng Xiao, Yee Whye Teh, Junxian He, and Manling Li. Internalizing world models via self-play finetuning for agentic rl. arXiv preprint arXiv:2510.15047, 2025. — [arXiv:2510.15047](https://arxiv.org/abs/2510.15047) [find](https://scholar.google.com/scholar?q=Shiqi+Chen%2C+Tongyao+Zhu%2C+Zian+Wang%2C+Jinghan+Zhang%2C+Kangrui+Wang%2C+Siyang+Gao%2C+Teng+Xiao%2C+Yee+Whye+Teh%2C+Junxian+He%2C+and+Manling+Li%2E+Internalizing+world+models+via+self%2Dplay+finetuning+for+agentic+rl%2E+ar)
16. Jiali Cheng, Anjishnu Kumar, Roshan Lal, Rishi Rajasekaran, Hani Ramezani, Omar Zia Khan, Oleg Rokhlenko, Sunny Chiu-Webster, Gang Hua, and Hadi Amiri. Webatlas: An llm agent with experience-driven memory and action simulation. arXiv preprint arXiv:2510.22732, 2025. — [arXiv:2510.22732](https://arxiv.org/abs/2510.22732) [find](https://scholar.google.com/scholar?q=Jiali+Cheng%2C+Anjishnu+Kumar%2C+Roshan+Lal%2C+Rishi+Rajasekaran%2C+Hani+Ramezani%2C+Omar+Zia+Khan%2C+Oleg+Rokhlenko%2C+Sunny+Chiu%2DWebster%2C+Gang+Hua%2C+and+Hadi+Amiri%2E+Webatlas%3A+An+llm+agent+with+experience%2Ddriven+me)
17. Meng Chu, Xuan Billy Zhang, Kevin Qinghong Lin, Lingdong Kong, Jize Zhang, Teng Tu, Weijian Ma, Ziqi Huang, Senqiao Yang, Wei Huang, et al. Agentic world modeling: Foundations, capabilities, laws, and beyond. arXiv preprint arXiv:2604.22748, 2026. — [arXiv:2604.22748](https://arxiv.org/abs/2604.22748) [find](https://scholar.google.com/scholar?q=Meng+Chu%2C+Xuan+Billy+Zhang%2C+Kevin+Qinghong+Lin%2C+Lingdong+Kong%2C+Jize+Zhang%2C+Teng+Tu%2C+Weijian+Ma%2C+Ziqi+Huang%2C+Senqiao+Yang%2C+Wei+Huang%2C+et+al%2E+Agentic+world+modeling%3A+Foundations%2C+capabilities%2C+laws%2C+and)
18. Santiago Cifuentes. General agents contain world models, even under partial observability and stochasticity. arXiv preprint arXiv:2602.03146, 2026. — [arXiv:2602.03146](https://arxiv.org/abs/2602.03146) [find](https://scholar.google.com/scholar?q=Santiago+Cifuentes%2E+General+agents+contain+world+models%2C+even+under+partial+observability+and+stochasticity%2E+arXiv+preprint+arXiv%3A2602%2E03146%2C+2026%2E+%E2%80%94+%5BarXiv%3A2602%2E03146%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2602%2E0314)
19. Jade Copet, Quentin Carbonneaux, Gal Cohen, Jonas Gehring, Jacob Kahn, Jannik Kossen, Felix Kreuk, Emily McMilin, Michel Meyer, Yuxiang Wei, et al. Cwm: An open-weights llm for research on code generation with world models. arXiv preprint arXiv:2510.02387, 2025. — [arXiv:2510.02387](https://arxiv.org/abs/2510.02387) [find](https://scholar.google.com/scholar?q=Jade+Copet%2C+Quentin+Carbonneaux%2C+Gal+Cohen%2C+Jonas+Gehring%2C+Jacob+Kahn%2C+Jannik+Kossen%2C+Felix+Kreuk%2C+Emily+McMilin%2C+Michel+Meyer%2C+Yuxiang+Wei%2C+et+al%2E+Cwm%3A+An+open%2Dweights+llm+for+research+on+code+genera)
20. Google DeepMind. Gemini 3 flash model card, 2025. URL [https://storage.googleapis.com/deepmin](https://storage.googleapis.com/deepmin) d-media/Model-Cards/Gemini-3-Flash-Model-Card.pdf. [↗](https://storage.googleapis.com/deepmin)
21. Google DeepMind. Gemini 3.1 pro model card, 2026. URL [https://storage.googleapis.com/deepmin](https://storage.googleapis.com/deepmin) d-media/Model-Cards/Gemini-3-1-Pro-Model-Card.pdf. [↗](https://storage.googleapis.com/deepmin)
22. DeepSeek-AI. [Deepseek-v4: Towards highly efficient million-token context intelligence](https://md.research.elamur.ai/free/en/papers/deepseek-ai-2026-deepseek-v4-highly-efficient/), 2026.
23. Florent Delgrange. Foundation world models for agents that learn, verify, and adapt reliably beyond static environments. arXiv preprint arXiv:2602.23997, 2026. — [arXiv:2602.23997](https://arxiv.org/abs/2602.23997) [find](https://scholar.google.com/scholar?q=Florent+Delgrange%2E+Foundation+world+models+for+agents+that+learn%2C+verify%2C+and+adapt+reliably+beyond+static+environments%2E+arXiv+preprint+arXiv%3A2602%2E23997%2C+2026%2E+%E2%80%94+%5BarXiv%3A2602%2E23997%5D%28https%3A%2F%2Farxiv%2Eorg%2Fa)
24. Xiang Deng, Jeff Da, Edwin Pan, Yannis Yiming He, Charles Ide, Kanak Garg, Niklas Lauffer, Andrew Park, Nitin Pasari, Chetan Rane, et al. Swe-bench pro: Can ai agents solve long-horizon software engineering tasks? arXiv preprint arXiv:2509.16941, 2025. — [arXiv:2509.16941](https://arxiv.org/abs/2509.16941) [find](https://scholar.google.com/scholar?q=Xiang+Deng%2C+Jeff+Da%2C+Edwin+Pan%2C+Yannis+Yiming+He%2C+Charles+Ide%2C+Kanak+Garg%2C+Niklas+Lauffer%2C+Andrew+Park%2C+Nitin+Pasari%2C+Chetan+Rane%2C+et+al%2E+Swe%2Dbench+pro%3A+Can+ai+agents+solve+long%2Dhorizon+software+engin)
25. Hang Ding, Peidong Liu, Junqiao Wang, Ziwei Ji, Meng Cao, Rongzhao Zhang, Lynn Ai, Eric Yang, Tianyu Shi, and Lei Yu. Dynaweb: Model-based reinforcement learning of web agents. arXiv preprint arXiv:2601.22149, 2026. — [arXiv:2601.22149](https://arxiv.org/abs/2601.22149) [find](https://scholar.google.com/scholar?q=Hang+Ding%2C+Peidong+Liu%2C+Junqiao+Wang%2C+Ziwei+Ji%2C+Meng+Cao%2C+Rongzhao+Zhang%2C+Lynn+Ai%2C+Eric+Yang%2C+Tianyu+Shi%2C+and+Lei+Yu%2E+Dynaweb%3A+Model%2Dbased+reinforcement+learning+of+web+agents%2E+arXiv+preprint+arXiv%3A26)
26. Guanting Dong, Junting Lu, Junjie Huang, Wanjun Zhong, Longxiang Liu, Shijue Huang, Zhenyu Li, Yang Zhao, Xiaoshuai Song, Xiaoxi Li, et al. Agent-world: Scaling real-world environment synthesis for evolving general agent intelligence. arXiv preprint arXiv:2604.18292, 2026. — [arXiv:2604.18292](https://arxiv.org/abs/2604.18292) [find](https://scholar.google.com/scholar?q=Guanting+Dong%2C+Junting+Lu%2C+Junjie+Huang%2C+Wanjun+Zhong%2C+Longxiang+Liu%2C+Shijue+Huang%2C+Zhenyu+Li%2C+Yang+Zhao%2C+Xiaoshuai+Song%2C+Xiaoxi+Li%2C+et+al%2E+Agent%2Dworld%3A+Scaling+real%2Dworld+environment+synthesis+for+ev)
27. Sicheng Fan, Qingyun Shi, Shengze Xu, Shengbo Cai, Tieyong Zeng, Li Ling, Yanyi Shang, and Dehan Kong. Webfactory: Automated compression of foundational language intelligence into grounded web agents. arXiv preprint arXiv:2603.05044, 2026. — [arXiv:2603.05044](https://arxiv.org/abs/2603.05044) [find](https://scholar.google.com/scholar?q=Sicheng+Fan%2C+Qingyun+Shi%2C+Shengze+Xu%2C+Shengbo+Cai%2C+Tieyong+Zeng%2C+Li+Ling%2C+Yanyi+Shang%2C+and+Dehan+Kong%2E+Webfactory%3A+Automated+compression+of+foundational+language+intelligence+into+grounded+web+agents%2E)
28. Yuchen Fan, Kaiyan Zhang, Heng Zhou, Yuxin Zuo, Yanxu Chen, Yu Fu, Xinwei Long, Xuekai Zhu, Che Jiang, Yuchen Zhang, et al. Ssrl: Self-search reinforcement learning. arXiv preprint arXiv:2508.10874, 2025. — [arXiv:2508.10874](https://arxiv.org/abs/2508.10874) [find](https://scholar.google.com/scholar?q=Yuchen+Fan%2C+Kaiyan+Zhang%2C+Heng+Zhou%2C+Yuxin+Zuo%2C+Yanxu+Chen%2C+Yu+Fu%2C+Xinwei+Long%2C+Xuekai+Zhu%2C+Che+Jiang%2C+Yuchen+Zhang%2C+et+al%2E+Ssrl%3A+Self%2Dsearch+reinforcement+learning%2E+arXiv+preprint+arXiv%3A2508%2E10874%2C+2)
29. Runnan Fang, Shihao Cai, Baixuan Li, Jialong Wu, Guangyu Li, Wenbiao Yin, Xinyu Wang, Xiaobin Wang, Liangcai Su, Zhen Zhang, et al. Towards general agentic intelligence via environment scaling. arXiv preprint arXiv:2509.13311, 2025. [find](https://scholar.google.com/scholar?q=Runnan+Fang%2C+Shihao+Cai%2C+Baixuan+Li%2C+Jialong+Wu%2C+Guangyu+Li%2C+Wenbiao+Yin%2C+Xinyu+Wang%2C+Xiaobin+Wang%2C+Liangcai+Su%2C+Zhen+Zhang%2C+et+al%2E+Towards+general+agentic+intelligence+via+environment+scaling%2E+arXiv+)
30. Dayuan Fu, Shenyu Wu, Yunze Wu, Zerui Peng, Yaxing Huang, Jie Sun, Ji Zeng, Mohan Jiang, Lin Zhang, Yukun Li, et al. davinci-env: Open swe environment synthesis at scale. arXiv preprint arXiv:2603.13023, 2026. — [arXiv:2603.13023](https://arxiv.org/abs/2603.13023) [find](https://scholar.google.com/scholar?q=Dayuan+Fu%2C+Shenyu+Wu%2C+Yunze+Wu%2C+Zerui+Peng%2C+Yaxing+Huang%2C+Jie+Sun%2C+Ji+Zeng%2C+Mohan+Jiang%2C+Lin+Zhang%2C+Yukun+Li%2C+et+al%2E+davinci%2Denv%3A+Open+swe+environment+synthesis+at+scale%2E+arXiv+preprint+arXiv%3A2603%2E130)
31. Kanishk Gandhi, Shivam Garg, Noah D Goodman, and Dimitris Papailiopoulos. Endless terminals: Scaling rl environments for terminal agents. arXiv preprint arXiv:2601.16443, 2026. — [arXiv:2601.16443](https://arxiv.org/abs/2601.16443) [find](https://scholar.google.com/scholar?q=Kanishk+Gandhi%2C+Shivam+Garg%2C+Noah+D+Goodman%2C+and+Dimitris+Papailiopoulos%2E+Endless+terminals%3A+Scaling+rl+environments+for+terminal+agents%2E+arXiv+preprint+arXiv%3A2601%2E16443%2C+2026%2E+%E2%80%94+%5BarXiv%3A2601%2E16443%5D%28ht)
32. Yifei Gao, Junhong Ye, Jiaqi Wang, and Jitao Sang. Websynthesis: World-model-guided mcts for efficient webui-trajectory synthesis. arXiv preprint arXiv:2507.04370, 2025. — [arXiv:2507.04370](https://arxiv.org/abs/2507.04370) [find](https://scholar.google.com/scholar?q=Yifei+Gao%2C+Junhong+Ye%2C+Jiaqi+Wang%2C+and+Jitao+Sang%2E+Websynthesis%3A+World%2Dmodel%2Dguided+mcts+for+efficient+webui%2Dtrajectory+synthesis%2E+arXiv+preprint+arXiv%3A2507%2E04370%2C+2025%2E+%E2%80%94+%5BarXiv%3A2507%2E04370%5D%28https%3A%2F%2Fa)
33. Yu Gu, Kai Zhang, Yuting Ning, Boyuan Zheng, Boyu Gou, Tianci Xue, Cheng Chang, Sanjari Srivastava, Yanan Xie, Peng Qi, et al. Is your llm secretly a world model of the internet? model-based planning for web agents. arXiv preprint arXiv:2411.06559, 2024. — [arXiv:2411.06559](https://arxiv.org/abs/2411.06559) [find](https://scholar.google.com/scholar?q=Yu+Gu%2C+Kai+Zhang%2C+Yuting+Ning%2C+Boyuan+Zheng%2C+Boyu+Gou%2C+Tianci+Xue%2C+Cheng+Chang%2C+Sanjari+Srivastava%2C+Yanan+Xie%2C+Peng+Qi%2C+et+al%2E+Is+your+llm+secretly+a+world+model+of+the+internet%3F+model%2Dbased+planning+)
34. Yiming Guan, Rui Yu, John Zhang, Lu Wang, Chaoyun Zhang, Liqun Li, Bo Qiao, Si Qin, He Huang, Fangkai Yang, et al. Computer-using world model. arXiv preprint arXiv:2602.17365, 2026. — [arXiv:2602.17365](https://arxiv.org/abs/2602.17365) [find](https://scholar.google.com/scholar?q=Yiming+Guan%2C+Rui+Yu%2C+John+Zhang%2C+Lu+Wang%2C+Chaoyun+Zhang%2C+Liqun+Li%2C+Bo+Qiao%2C+Si+Qin%2C+He+Huang%2C+Fangkai+Yang%2C+et+al%2E+Computer%2Dusing+world+model%2E+arXiv+preprint+arXiv%3A2602%2E17365%2C+2026%2E+%E2%80%94+%5BarXiv%3A2602%2E1736)
35. Anisha Gunjal, Anthony Wang, Elaine Lau, Vaskar Nath, Yunzhong He, Bing Liu, and Sean Hendryx. Rubrics as rewards: Reinforcement learning beyond verifiable domains. arXiv preprint arXiv:2507.17746, 2025. [find](https://scholar.google.com/scholar?q=Anisha+Gunjal%2C+Anthony+Wang%2C+Elaine+Lau%2C+Vaskar+Nath%2C+Yunzhong+He%2C+Bing+Liu%2C+and+Sean+Hendryx%2E+Rubrics+as+rewards%3A+Reinforcement+learning+beyond+verifiable+domains%2E+arXiv+preprint+arXiv%3A2507%2E17746%2C+20)
36. Jiacheng Guo, Ling Yang, Peter Chen, Qixin Xiao, Yinjie Wang, Xinzhe Juan, Jiahao Qiu, Ke Shen, and Mengdi Wang. Genenv: Difficulty-aligned co-evolution between llm agents and environment simulators. arXiv preprint arXiv:2512.19682, 2025a. — [arXiv:2512.19682](https://arxiv.org/abs/2512.19682) [find](https://scholar.google.com/scholar?q=Jiacheng+Guo%2C+Ling+Yang%2C+Peter+Chen%2C+Qixin+Xiao%2C+Yinjie+Wang%2C+Xinzhe+Juan%2C+Jiahao+Qiu%2C+Ke+Shen%2C+and+Mengdi+Wang%2E+Genenv%3A+Difficulty%2Daligned+co%2Devolution+between+llm+agents+and+environment+simulators%2E+)
37. Shangmin Guo, Omar Darwiche Domingues, Raphaël Avalos, Aaron Courville, and Florian Strub. World modelling improves language model agents. arXiv preprint arXiv:2506.02918, 2025b. — [arXiv:2506.02918](https://arxiv.org/abs/2506.02918) [find](https://scholar.google.com/scholar?q=Shangmin+Guo%2C+Omar+Darwiche+Domingues%2C+Rapha%C3%ABl+Avalos%2C+Aaron+Courville%2C+and+Florian+Strub%2E+World+modelling+improves+language+model+agents%2E+arXiv+preprint+arXiv%3A2506%2E02918%2C+2025b%2E+%E2%80%94+%5BarXiv%3A2506%2E02918%5D%28)
38. Danijar Hafner, Jurgis Pasukonis, Jimmy Ba, and Timothy Lillicrap. Mastering diverse domains through world models. arXiv preprint arXiv:2301.04104, 2023. [find](https://scholar.google.com/scholar?q=Danijar+Hafner%2C+Jurgis+Pasukonis%2C+Jimmy+Ba%2C+and+Timothy+Lillicrap%2E+Mastering+diverse+domains+through+world+models%2E+arXiv+preprint+arXiv%3A2301%2E04104%2C+2023%2E)
39. Danijar Hafner, Wilson Yan, and Timothy Lillicrap. Training agents inside of scalable world models. arXiv preprint arXiv:2509.24527, 2025. [find](https://scholar.google.com/scholar?q=Danijar+Hafner%2C+Wilson+Yan%2C+and+Timothy+Lillicrap%2E+Training+agents+inside+of+scalable+world+models%2E+arXiv+preprint+arXiv%3A2509%2E24527%2C+2025%2E)
40. Tiantian He, Yihang Chen, Keyue Jiang, Ka Yiu Lee, Kaiwen Zhou, Kun Shao, and Shuai Wang. Ee-mcp: Self-evolving mcp-gui agents via automated environment generation and experience learning. arXiv preprint arXiv:2604.09815, 2026. — [arXiv:2604.09815](https://arxiv.org/abs/2604.09815) [find](https://scholar.google.com/scholar?q=Tiantian+He%2C+Yihang+Chen%2C+Keyue+Jiang%2C+Ka+Yiu+Lee%2C+Kaiwen+Zhou%2C+Kun+Shao%2C+and+Shuai+Wang%2E+Ee%2Dmcp%3A+Self%2Devolving+mcp%2Dgui+agents+via+automated+environment+generation+and+experience+learning%2E+arXiv+prepr)
41. Bohan Hou, Gen Li, Jindou Jia, Tuo An, Xinying Guo, Sicong Leng, Haoran Geng, Yanjie Ze, Tatsuya Harada, Philip Torr, et al. World model for robot learning: A comprehensive survey. arXiv preprint arXiv:2605.00080, 2026. — [arXiv:2605.00080](https://arxiv.org/abs/2605.00080) [find](https://scholar.google.com/scholar?q=Bohan+Hou%2C+Gen+Li%2C+Jindou+Jia%2C+Tuo+An%2C+Xinying+Guo%2C+Sicong+Leng%2C+Haoran+Geng%2C+Yanjie+Ze%2C+Tatsuya+Harada%2C+Philip+Torr%2C+et+al%2E+World+model+for+robot+learning%3A+A+comprehensive+survey%2E+arXiv+preprint+arXi)
42. Mengkang Hu, Tianxing Chen, Yude Zou, Yuheng Lei, Qiguang Chen, Ming Li, Yao Mu, Hongyuan Zhang, Wenqi Shao, and Ping Luo. Text2world: Benchmarking large language models for symbolic world model generation. In Findings of the Association for Computational Linguistics: ACL 2025, pp. 26043–26066, 2025… [find](https://scholar.google.com/scholar?q=Mengkang+Hu%2C+Tianxing+Chen%2C+Yude+Zou%2C+Yuheng+Lei%2C+Qiguang+Chen%2C+Ming+Li%2C+Yao+Mu%2C+Hongyuan+Zhang%2C+Wenqi+Shao%2C+and+Ping+Luo%2E+Text2world%3A+Benchmarking+large+language+models+for+symbolic+world+model+gener)
43. Mengkang Hu, Bowei Xia, Yuran Wu, Ailing Yu, Yude Zou, Qiguang Chen, Shijian Wang, Jiarui Jin, Kexin Li, Wenxiang Jiao, et al. Agent2world: Learning to generate symbolic world models via adaptive multi-agent feedback. arXiv preprint arXiv:2512.22336, 2025b. — [arXiv:2512.22336](https://arxiv.org/abs/2512.22336) [find](https://scholar.google.com/scholar?q=Mengkang+Hu%2C+Bowei+Xia%2C+Yuran+Wu%2C+Ailing+Yu%2C+Yude+Zou%2C+Qiguang+Chen%2C+Shijian+Wang%2C+Jiarui+Jin%2C+Kexin+Li%2C+Wenxiang+Jiao%2C+et+al%2E+Agent2world%3A+Learning+to+generate+symbolic+world+models+via+adaptive+mult)
44. Siqiao Huang, Jialong Wu, Qixing Zhou, Shangchen Miao, and Mingsheng Long. Vid2world: Crafting video diffusion models to interactive world models. arXiv preprint arXiv:2505.14357, 2025a. — [arXiv:2505.14357](https://arxiv.org/abs/2505.14357) [find](https://scholar.google.com/scholar?q=Siqiao+Huang%2C+Jialong+Wu%2C+Qixing+Zhou%2C+Shangchen+Miao%2C+and+Mingsheng+Long%2E+Vid2world%3A+Crafting+video+diffusion+models+to+interactive+world+models%2E+arXiv+preprint+arXiv%3A2505%2E14357%2C+2025a%2E+%E2%80%94+%5BarXiv%3A2505)
45. Youling Huang, Guanqiao Chen, Junchi Yao, Lu Wang, Fangkai Yang, Chao Du, ChenZhuo Zhao, Pu Zhao, Qingwei Lin, Saravan Rajmohan, et al. Beyond state consistency: Behavior consistency in text-based world models. arXiv preprint arXiv:2604.13824, 2026. — [arXiv:2604.13824](https://arxiv.org/abs/2604.13824) [find](https://scholar.google.com/scholar?q=Youling+Huang%2C+Guanqiao+Chen%2C+Junchi+Yao%2C+Lu+Wang%2C+Fangkai+Yang%2C+Chao+Du%2C+ChenZhuo+Zhao%2C+Pu+Zhao%2C+Qingwei+Lin%2C+Saravan+Rajmohan%2C+et+al%2E+Beyond+state+consistency%3A+Behavior+consistency+in+text%2Dbased+wor)
46. Yuchen Huang, Sijia Li, Minghao Liu, Wei Liu, Shijue Huang, Zhiyuan Fan, Hou Pong Chan, and Yi R Fung. Environment scaling for interactive agentic experience collection: A survey. arXiv preprint arXiv:2511.09586, 2025b. — [arXiv:2511.09586](https://arxiv.org/abs/2511.09586) [find](https://scholar.google.com/scholar?q=Yuchen+Huang%2C+Sijia+Li%2C+Minghao+Liu%2C+Wei+Liu%2C+Shijue+Huang%2C+Zhiyuan+Fan%2C+Hou+Pong+Chan%2C+and+Yi+R+Fung%2E+Environment+scaling+for+interactive+agentic+experience+collection%3A+A+survey%2E+arXiv+preprint+arXiv)
47. Carlos E Jimenez, John Yang, Alexander Wettig, Shunyu Yao, Kexin Pei, Ofir Press, and Karthik Narasimhan. [Swe-bench: Can language models resolve real-world github issues?](https://md.research.elamur.ai/free/en/papers/jimenez-2024-swe-bench-can-language-models/) In International Conference on Learning Representations, volume 2024, pp. 54107–54157, 2024.
48. Oğuzhan Fatih Kar, Roman Bachmann, Yuanzheng Gong, Anders Boesen Lindbo Larsen, and Afshin Dehghan. Weblica: Scalable and reproducible training environments for visual web agents, 2026. URL https://arxiv.org/abs/2605.06761. — [arXiv:2605.06761](https://arxiv.org/abs/2605.06761) [↗](https://arxiv.org/abs/2605.06761)
49. Andrej Karpathy. autoresearch. [https://github.com/karpathy/autoresearch](https://github.com/karpathy/autoresearch), 2026. [↗](https://github.com/karpathy/autoresearch)
50. Woosung Koh, Sungjun Han, Segyu Lee, Se-Young Yun, and Jamin Shin. Generative visual code mobile world models. arXiv preprint arXiv:2602.01576, 2026. — [arXiv:2602.01576](https://arxiv.org/abs/2602.01576) [find](https://scholar.google.com/scholar?q=Woosung+Koh%2C+Sungjun+Han%2C+Segyu+Lee%2C+Se%2DYoung+Yun%2C+and+Jamin+Shin%2E+Generative+visual+code+mobile+world+models%2E+arXiv+preprint+arXiv%3A2602%2E01576%2C+2026%2E+%E2%80%94+%5BarXiv%3A2602%2E01576%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2602%2E01)
51. Yann LeCun et al. A path towards autonomous machine intelligence version 0.9. 2, 2022-06-27. Open Review, 62(1):1–62, 2022. [find](https://scholar.google.com/scholar?q=Yann+LeCun+et+al%2E+A+path+towards+autonomous+machine+intelligence+version+0%2E9%2E+2%2C+2022%2D06%2D27%2E+Open+Review%2C+62%281%29%3A1%E2%80%9362%2C+2022%2E)
52. Wolfgang Lehrach, Daniel Hennes, Miguel Lazaro-Gredilla, Xinghua Lou, Carter Wendelken, Zun Li, Antoine Dedieu, Jordi Grau-Moya, Marc Lanctot, Atil Iscen, et al. Code world models for general game playing. arXiv preprint arXiv:2510.04542, 2025. — [arXiv:2510.04542](https://arxiv.org/abs/2510.04542) [find](https://scholar.google.com/scholar?q=Wolfgang+Lehrach%2C+Daniel+Hennes%2C+Miguel+Lazaro%2DGredilla%2C+Xinghua+Lou%2C+Carter+Wendelken%2C+Zun+Li%2C+Antoine+Dedieu%2C+Jordi+Grau%2DMoya%2C+Marc+Lanctot%2C+Atil+Iscen%2C+et+al%2E+Code+world+models+for+general+game+pla)
53. Guillaume Levy, Cédric Colas, Pierre-Yves Oudeyer, Thomas Carta, and Clément Romac. Worldllm: Improving llms’ world modeling using curiosity-driven theory-making. arXiv preprint arXiv:2506.06725, 2025. — [arXiv:2506.06725](https://arxiv.org/abs/2506.06725) [find](https://scholar.google.com/scholar?q=Guillaume+Levy%2C+C%C3%A9dric+Colas%2C+Pierre%2DYves+Oudeyer%2C+Thomas+Carta%2C+and+Cl%C3%A9ment+Romac%2E+Worldllm%3A+Improving+llms%E2%80%99+world+modeling+using+curiosity%2Ddriven+theory%2Dmaking%2E+arXiv+preprint+arXiv%3A2506%2E06725%2C+2025)
54. Junlong Li, Wenshuo Zhao, Jian Zhao, Weihao Zeng, Haoze Wu, Xiaochen Wang, Rui Ge, Yuxuan Cao, Yuzhen Huang, Wei Liu, et al. [The tool decathlon: Benchmarking language agents for diverse, realistic, and long-horizon task execution](https://md.research.elamur.ai/free/en/papers/li-2025-tool-decathlon-benchmarking-language-agents/). arXiv preprint arXiv:2510.25726, 2025a.
55. Wanli Li, Bince Qu, Bo Pan, Jianyu Zhang, Zheng Liu, Pan Zhang, Wei Chen, and Bo Zhang. Literesearcher: A scalable agentic rl training framework for deep research agent. arXiv preprint arXiv:2604.17931, 2026a. — [arXiv:2604.17931](https://arxiv.org/abs/2604.17931) [find](https://scholar.google.com/scholar?q=Wanli+Li%2C+Bince+Qu%2C+Bo+Pan%2C+Jianyu+Zhang%2C+Zheng+Liu%2C+Pan+Zhang%2C+Wei+Chen%2C+and+Bo+Zhang%2E+Literesearcher%3A+A+scalable+agentic+rl+training+framework+for+deep+research+agent%2E+arXiv+preprint+arXiv%3A2604%2E1793)
56. Xirui Li, Ming Li, Ion Stoica, Cho-Jui Hsieh, and Tianyi Zhou. Clawenvkit: Automatic environment generation for claw-like agents. arXiv preprint arXiv:2604.18543, 2026b. — [arXiv:2604.18543](https://arxiv.org/abs/2604.18543) [find](https://scholar.google.com/scholar?q=Xirui+Li%2C+Ming+Li%2C+Ion+Stoica%2C+Cho%2DJui+Hsieh%2C+and+Tianyi+Zhou%2E+Clawenvkit%3A+Automatic+environment+generation+for+claw%2Dlike+agents%2E+arXiv+preprint+arXiv%3A2604%2E18543%2C+2026b%2E+%E2%80%94+%5BarXiv%3A2604%2E18543%5D%28https%3A%2F%2Fa)
57. Yixia Li, Hongru Wang, Jiahao Qiu, Zhenfei Yin, Dongdong Zhang, Cheng Qian, Zeping Li, Pony Ma, Guanhua Chen, and Heng Ji. From word to world: Can large language models be implicit text-based world models? arXiv preprint arXiv:2512.18832, 2025b. [find](https://scholar.google.com/scholar?q=Yixia+Li%2C+Hongru+Wang%2C+Jiahao+Qiu%2C+Zhenfei+Yin%2C+Dongdong+Zhang%2C+Cheng+Qian%2C+Zeping+Li%2C+Pony+Ma%2C+Guanhua+Chen%2C+and+Heng+Ji%2E+From+word+to+world%3A+Can+large+language+models+be+implicit+text%2Dbased+world+mo)
58. Yixia Li, Hongru Wang, Peng Lai, Zhiwen Ruan, He Zhu, Youxin Zhu, Ganlong Zhao, Minda Hu, Yun Chen, Sibei Yang, Peng Li, Jeff Z. Pan, Jia Pan, Guanhua Chen, Yang Liu, and Guanbin Li. Bridging the agent-world gap: Text world models for llm-based agents, 2026c. URL https://arxiv.org/abs/2606 .09032. [↗](https://arxiv.org/abs/2606)
59. Yuetai Li, Huseyin A Inan, Xiang Yue, Wei-Ning Chen, Lukas Wutschitz, Janardhan Kulkarni, Radha Poovendran, Robert Sim, and Saravan Rajmohan. Simulating environments with reasoning models for agent training. arXiv preprint arXiv:2511.01824, 2025c. — [arXiv:2511.01824](https://arxiv.org/abs/2511.01824) [find](https://scholar.google.com/scholar?q=Yuetai+Li%2C+Huseyin+A+Inan%2C+Xiang+Yue%2C+Wei%2DNing+Chen%2C+Lukas+Wutschitz%2C+Janardhan+Kulkarni%2C+Radha+Poovendran%2C+Robert+Sim%2C+and+Saravan+Rajmohan%2E+Simulating+environments+with+reasoning+models+for+agent+tr)
60. Youwei Liu, Jian Wang, Hanlin Wang, Beichen Guo, and Wenjie Li. Imagine-then-plan: Agent learning from adaptive lookahead with world models. arXiv preprint arXiv:2601.08955, 2026. — [arXiv:2601.08955](https://arxiv.org/abs/2601.08955) [find](https://scholar.google.com/scholar?q=Youwei+Liu%2C+Jian+Wang%2C+Hanlin+Wang%2C+Beichen+Guo%2C+and+Wenjie+Li%2E+Imagine%2Dthen%2Dplan%3A+Agent+learning+from+adaptive+lookahead+with+world+models%2E+arXiv+preprint+arXiv%3A2601%2E08955%2C+2026%2E+%E2%80%94+%5BarXiv%3A2601%2E08955%5D)
61. Mike A Merrill, Alexander G Shaw, Nicholas Carlini, Boxuan Li, Harsh Raj, Ivan Bercovich, Lin Shi, Jeong Yeon Shin, Thomas Walshe, E Kelly Buchanan, et al. Terminal-bench: Benchmarking agents on hard, realistic tasks in command line interfaces. arXiv preprint arXiv:2601.11868, 2026. — [arXiv:2601.11868](https://arxiv.org/abs/2601.11868) [find](https://scholar.google.com/scholar?q=Mike+A+Merrill%2C+Alexander+G+Shaw%2C+Nicholas+Carlini%2C+Boxuan+Li%2C+Harsh+Raj%2C+Ivan+Bercovich%2C+Lin+Shi%2C+Jeong+Yeon+Shin%2C+Thomas+Walshe%2C+E+Kelly+Buchanan%2C+et+al%2E+Terminal%2Dbench%3A+Benchmarking+agents+on+hard%2C)
62. Vincent Micheli, Eloi Alonso, and François Fleuret. Transformers are sample-efficient world models. arXiv preprint arXiv:2209.00588, 2022. — [arXiv:2209.00588](https://arxiv.org/abs/2209.00588) [find](https://scholar.google.com/scholar?q=Vincent+Micheli%2C+Eloi+Alonso%2C+and+Fran%C3%A7ois+Fleuret%2E+Transformers+are+sample%2Defficient+world+models%2E+arXiv+preprint+arXiv%3A2209%2E00588%2C+2022%2E+%E2%80%94+%5BarXiv%3A2209%2E00588%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2209%2E00588%29)
63. AI MiniMax. Minimax m2. 7: Early echoes of self-evolution, 2026. [find](https://scholar.google.com/scholar?q=AI+MiniMax%2E+Minimax+m2%2E+7%3A+Early+echoes+of+self%2Devolution%2C+2026%2E)
64. OpenAI. Introducing gpt-5.2, 2025. URL [https://openai.com/index/introducing-gpt-5-2/](https://openai.com/index/introducing-gpt-5-2/). [↗](https://openai.com/index/introducing-gpt-5-2/)
65. OpenAI. Introducing gpt-5.4, 2026. URL [https://openai.com/index/introducing-gpt-5-4/](https://openai.com/index/introducing-gpt-5-4/). [↗](https://openai.com/index/introducing-gpt-5-4/)
66. OpenClaw. Openclaw. [https://github.com/openclaw/openclaw](https://github.com/openclaw/openclaw), 2026. Open-source personal AI assistant, version 2026.3.8, accessed 2026-03-09. [↗](https://github.com/openclaw/openclaw)
67. Shishir G Patil, Huanzhi Mao, Fanjia Yan, Charlie Cheng-Jie Ji, Vishnu Suresh, Ion Stoica, and Joseph E Gonzalez. The berkeley function calling leaderboard (bfcl): From tool use to agentic evaluation of large language models. In Forty-second International Conference on Machine Learning, 2025. [find](https://scholar.google.com/scholar?q=Shishir+G+Patil%2C+Huanzhi+Mao%2C+Fanjia+Yan%2C+Charlie+Cheng%2DJie+Ji%2C+Vishnu+Suresh%2C+Ion+Stoica%2C+and+Joseph+E+Gonzalez%2E+The+berkeley+function+calling+leaderboard+%28bfcl%29%3A+From+tool+use+to+agentic+evaluation+)
68. Tejal Patwardhan, Rachel Dias, Elizabeth Proehl, Grace Kim, Michele Wang, Olivia Watkins, Simón Posada Fishman, Marwan Aljubeh, Phoebe Thacker, Laurance Fauconnet, et al. Gdpval: Evaluating ai model performance on real-world economically valuable tasks. arXiv preprint arXiv:2510.04374, 2025. [find](https://scholar.google.com/scholar?q=Tejal+Patwardhan%2C+Rachel+Dias%2C+Elizabeth+Proehl%2C+Grace+Kim%2C+Michele+Wang%2C+Olivia+Watkins%2C+Sim%C3%B3n+Posada+Fishman%2C+Marwan+Aljubeh%2C+Phoebe+Thacker%2C+Laurance+Fauconnet%2C+et+al%2E+Gdpval%3A+Evaluating+ai+model+p)
69. Cheng Qian, Emre Can Acikgoz, Bingxuan Li, Xiusi Chen, Yuji Zhang, Bingxiang He, Qinyu Luo, Dilek Hakkani-Tür, Gokhan Tur, Yunzhu Li, et al. [Current agents fail to leverage world model as tool for foresight](https://md.research.elamur.ai/free/en/papers/qian-2026-current-agents-fail-leverage-world/). arXiv preprint arXiv:2601.03905, 2026.
70. Yifu Qiu, Zheng Zhao, Waylon Li, Yftah Ziser, Anna Korhonen, Shay B Cohen, and Edoardo M Ponti. Self-improving world modelling with latent actions. arXiv preprint arXiv:2602.06130, 2026. — [arXiv:2602.06130](https://arxiv.org/abs/2602.06130) [find](https://scholar.google.com/scholar?q=Yifu+Qiu%2C+Zheng+Zhao%2C+Waylon+Li%2C+Yftah+Ziser%2C+Anna+Korhonen%2C+Shay+B+Cohen%2C+and+Edoardo+M+Ponti%2E+Self%2Dimproving+world+modelling+with+latent+actions%2E+arXiv+preprint+arXiv%3A2602%2E06130%2C+2026%2E+%E2%80%94+%5BarXiv%3A2602)
71. Qwen Team. Qwen3.6-35B-A3B: Agentic coding power, now open to all, April 2026a. URL [https://qwen.ai/blog?id=qwen3.6-35b-a3b](https://qwen.ai/blog?id=qwen3.6-35b-a3b). [↗](https://qwen.ai/blog?id=qwen3.6-35b-a3b)
72. Qwen Team. Qwen3.6-Max-Preview: Smarter, sharper, still evolving, April 2026b. URL [https://qwen.a](https://qwen.a) i/blog?id=qwen3.6-max-preview. [↗](https://qwen.a)
73. Qwen Team. Qwen3.6-Plus: Towards real world agents, April 2026c. URL [https://qwen.ai/blog?id=qw](https://qwen.ai/blog?id=qw) en3.6. [↗](https://qwen.ai/blog?id=qw)
74. Babak Rahmani. Debugging code world models. arXiv preprint arXiv:2602.07672, 2026. — [arXiv:2602.07672](https://arxiv.org/abs/2602.07672) [find](https://scholar.google.com/scholar?q=Babak+Rahmani%2E+Debugging+code+world+models%2E+arXiv+preprint+arXiv%3A2602%2E07672%2C+2026%2E+%E2%80%94+%5BarXiv%3A2602%2E07672%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2602%2E07672%29)
75. Baochang Ren, Yunzhi Yao, Rui Sun, Shuofei Qiao, Ningyu Zhang, and Huajun Chen. Aligning agentic world models via knowledgeable experience learning. arXiv preprint arXiv:2601.13247, 2026. — [arXiv:2601.13247](https://arxiv.org/abs/2601.13247) [find](https://scholar.google.com/scholar?q=Baochang+Ren%2C+Yunzhi+Yao%2C+Rui+Sun%2C+Shuofei+Qiao%2C+Ningyu+Zhang%2C+and+Huajun+Chen%2E+Aligning+agentic+world+models+via+knowledgeable+experience+learning%2E+arXiv+preprint+arXiv%3A2601%2E13247%2C+2026%2E+%E2%80%94+%5BarXiv%3A260)
76. Jonathan Richens, David Abel, Alexis Bellot, and Tom Everitt. General agents contain world models. arXiv preprint arXiv:2506.01622, 2025. — [arXiv:2506.01622](https://arxiv.org/abs/2506.01622) [find](https://scholar.google.com/scholar?q=Jonathan+Richens%2C+David+Abel%2C+Alexis+Bellot%2C+and+Tom+Everitt%2E+General+agents+contain+world+models%2E+arXiv+preprint+arXiv%3A2506%2E01622%2C+2025%2E+%E2%80%94+%5BarXiv%3A2506%2E01622%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2506%2E01622%29)
77. William F Shen, Xinchi Qiu, Chenxi Whitehouse, Lisa Alazraki, Shashwat Goel, Francesco Barbieri, Timon Willi, Akhil Mathur, and Ilias Leontiadis. Rethinking rubric generation for improving llm judge and reward modeling for open-ended tasks. arXiv preprint arXiv:2602.05125, 2026a. [find](https://scholar.google.com/scholar?q=William+F+Shen%2C+Xinchi+Qiu%2C+Chenxi+Whitehouse%2C+Lisa+Alazraki%2C+Shashwat+Goel%2C+Francesco+Barbieri%2C+Timon+Willi%2C+Akhil+Mathur%2C+and+Ilias+Leontiadis%2E+Rethinking+rubric+generation+for+improving+llm+judge+a)
78. Zhouzhou Shen, Xueyu Hu, Xiyun Li, Tianqing Fang, Juncheng Li, and Shengyu Zhang. World-model-augmented web agents with action correction. arXiv preprint arXiv:2602.15384, 2026b. — [arXiv:2602.15384](https://arxiv.org/abs/2602.15384) [find](https://scholar.google.com/scholar?q=Zhouzhou+Shen%2C+Xueyu+Hu%2C+Xiyun+Li%2C+Tianqing+Fang%2C+Juncheng+Li%2C+and+Shengyu+Zhang%2E+World%2Dmodel%2Daugmented+web+agents+with+action+correction%2E+arXiv+preprint+arXiv%3A2602%2E15384%2C+2026b%2E+%E2%80%94+%5BarXiv%3A2602%2E15384%5D%28)
79. Vaishnavi Shrivastava, Piero Kauffmann, Ahmed Awadallah, and Dimitris Papailiopoulos. Echo: Terminal agents learn world models for free. arXiv preprint arXiv:2605.24517, 2026. — [arXiv:2605.24517](https://arxiv.org/abs/2605.24517) [find](https://scholar.google.com/scholar?q=Vaishnavi+Shrivastava%2C+Piero+Kauffmann%2C+Ahmed+Awadallah%2C+and+Dimitris+Papailiopoulos%2E+Echo%3A+Terminal+agents+learn+world+models+for+free%2E+arXiv+preprint+arXiv%3A2605%2E24517%2C+2026%2E+%E2%80%94+%5BarXiv%3A2605%2E24517%5D%28htt)
80. Xiaoshuai Song, Haofei Chang, Guanting Dong, Yutao Zhu, Ji-Rong Wen, and Zhicheng Dou. [Envscaler: Scaling tool-interactive environments for llm agent via programmatic synthesis](https://md.research.elamur.ai/free/en/papers/song-2026-envscaler-scaling-tool-interactive-environments/). arXiv preprint arXiv:2601.05808, 2026.
81. Liangcai Su, Zhen Zhang, Guangyu Li, Zhuo Chen, Chenxi Wang, Maojia Song, Xinyu Wang, Kuan Li, Jialong Wu, Xuanzhong Chen, et al. Scaling agents via continual pre-training. arXiv preprint arXiv:2509.13310, 2025. — [arXiv:2509.13310](https://arxiv.org/abs/2509.13310) [find](https://scholar.google.com/scholar?q=Liangcai+Su%2C+Zhen+Zhang%2C+Guangyu+Li%2C+Zhuo+Chen%2C+Chenxi+Wang%2C+Maojia+Song%2C+Xinyu+Wang%2C+Kuan+Li%2C+Jialong+Wu%2C+Xuanzhong+Chen%2C+et+al%2E+Scaling+agents+via+continual+pre%2Dtraining%2E+arXiv+preprint+arXiv%3A2509%2E1)
82. Hao Sun, Zile Qiao, Jiayan Guo, Xuanbo Fan, Yingyan Hou, Yong Jiang, Pengjun Xie, Yan Zhang, Fei Huang, and Jingren Zhou. [Zerosearch: Incentivize the search capability of llms without searching](https://md.research.elamur.ai/free/en/papers/sun-2025-zerosearch-incentivize-search-capability-llms/). arXiv preprint arXiv:2505.04588, 2025.
83. Shuang Sun, Huatong Song, Lisheng Huang, Jinhao Jiang, Ran Le, Zhihao Lv, Zongchao Chen, Yiwen Hu, Wenyang Luo, Wayne Xin Zhao, et al. Swe-world: Building software engineering agents in docker-free environments. arXiv preprint arXiv:2602.03419, 2026. — [arXiv:2602.03419](https://arxiv.org/abs/2602.03419) [find](https://scholar.google.com/scholar?q=Shuang+Sun%2C+Huatong+Song%2C+Lisheng+Huang%2C+Jinhao+Jiang%2C+Ran+Le%2C+Zhihao+Lv%2C+Zongchao+Chen%2C+Yiwen+Hu%2C+Wenyang+Luo%2C+Wayne+Xin+Zhao%2C+et+al%2E+Swe%2Dworld%3A+Building+software+engineering+agents+in+docker%2Dfree+en)
84. Chenming Tang, Hsiu-Yuan Huang, Weijie Liu, Junqiang Zheng, Saiyong Yang, and Yunfang Wu. Democratizing tool learning with environments fully simulated by a free 8b language model, 2026. URL https://arxiv.org/abs/2604.17739. — [arXiv:2604.17739](https://arxiv.org/abs/2604.17739) [↗](https://arxiv.org/abs/2604.17739)
85. Kimi Team, Tongtong Bai, Yifan Bai, Yiping Bao, SH Cai, Yuan Cao, Y Charles, HS Che, Cheng Chen, Guanduo Chen, et al. Kimi k2. 5: Visual agentic intelligence. arXiv preprint arXiv:2602.[02276, 2026](https://md.research.elamur.ai/free/en/papers/kimi-team-2026-kimi-k2-5-visual/).
86. Qwen Team. Qwen3.5: Accelerating productivity with native multimodal agents, February 2026. URL [https://qwen.ai/blog?id=qwen3.5](https://qwen.ai/blog?id=qwen3.5). [↗](https://qwen.ai/blog?id=qwen3.5)
87. Qwen Team and Alibaba Data. QwenClawBench: Real-user-distribution benchmark for openclaw agents, 2026. URL github.com/SKYLENAGE-AI/QwenClawBench. [find](https://scholar.google.com/scholar?q=Qwen+Team+and+Alibaba+Data%2E+QwenClawBench%3A+Real%2Duser%2Ddistribution+benchmark+for+openclaw+agents%2C+2026%2E+URL+github%2Ecom%2FSKYLENAGE%2DAI%2FQwenClawBench%2E)
88. Xiaoyu Tian, Haotian Wang, Shuaiting Chen, Hao Zhou, Kaichi Yu, Yudian Zhang, Jade Ouyang, Junxi Yin, Jiong Chen, Baoyan Guo, et al. Astra: Automated synthesis of agentic trajectories and reinforcement arenas. arXiv preprint arXiv:2601.21558, 2026. — [arXiv:2601.21558](https://arxiv.org/abs/2601.21558) [find](https://scholar.google.com/scholar?q=Xiaoyu+Tian%2C+Haotian+Wang%2C+Shuaiting+Chen%2C+Hao+Zhou%2C+Kaichi+Yu%2C+Yudian+Zhang%2C+Jade+Ouyang%2C+Junxi+Yin%2C+Jiong+Chen%2C+Baoyan+Guo%2C+et+al%2E+Astra%3A+Automated+synthesis+of+agentic+trajectories+and+reinforcemen)
89. Dunwei Tu, Hongyan Hao, Hansi Yang, Yihao Chen, Yi-Kai Zhang, Zhikang Xia, Yu Yang, Yueqing Sun, Xingchen Liu, Furao Shen, et al. Scaleenv: Scaling environment synthesis from scratch for generalist interactive tool-use agent training. arXiv preprint arXiv:2602.06820, 2026. — [arXiv:2602.06820](https://arxiv.org/abs/2602.06820) [find](https://scholar.google.com/scholar?q=Dunwei+Tu%2C+Hongyan+Hao%2C+Hansi+Yang%2C+Yihao+Chen%2C+Yi%2DKai+Zhang%2C+Zhikang+Xia%2C+Yu+Yang%2C+Yueqing+Sun%2C+Xingchen+Liu%2C+Furao+Shen%2C+et+al%2E+Scaleenv%3A+Scaling+environment+synthesis+from+scratch+for+generalist+in)
90. Dani Valevski, Yaniv Leviathan, Moab Arar, and Shlomi Fruchter. Diffusion models are real-time game engines. In International Conference on Learning Representations, volume 2025, pp. 73754–73776, 2025. [find](https://scholar.google.com/scholar?q=Dani+Valevski%2C+Yaniv+Leviathan%2C+Moab+Arar%2C+and+Shlomi+Fruchter%2E+Diffusion+models+are+real%2Dtime+game+engines%2E+In+International+Conference+on+Learning+Representations%2C+volume+2025%2C+pp%2E+73754%E2%80%9373776%2C+2025)
91. Kangrui Wang, Pingyue Zhang, Zihan Wang, Yaning Gao, Linjie Li, Qineng Wang, Hanyang Chen, Yiping Lu, Zhengyuan Yang, Lijuan Wang, et al. Vagen: Reinforcing world model reasoning for multi-turn vlm agents. Advances in Neural Information Processing Systems, 38:172871–172933, 2026a. [find](https://scholar.google.com/scholar?q=Kangrui+Wang%2C+Pingyue+Zhang%2C+Zihan+Wang%2C+Yaning+Gao%2C+Linjie+Li%2C+Qineng+Wang%2C+Hanyang+Chen%2C+Yiping+Lu%2C+Zhengyuan+Yang%2C+Lijuan+Wang%2C+et+al%2E+Vagen%3A+Reinforcing+world+model+reasoning+for+multi%2Dturn+vlm+ag)
92. Ruoyao Wang, Graham Todd, Ziang Xiao, Xingdi Yuan, Marc-Alexandre Côté, Peter Clark, and Peter Jansen. Can language models serve as text-based world simulators? In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pp. 1–17, 2024. [find](https://scholar.google.com/scholar?q=Ruoyao+Wang%2C+Graham+Todd%2C+Ziang+Xiao%2C+Xingdi+Yuan%2C+Marc%2DAlexandre+C%C3%B4t%C3%A9%2C+Peter+Clark%2C+and+Peter+Jansen%2E+Can+language+models+serve+as+text%2Dbased+world+simulators%3F+In+Proceedings+of+the+62nd+Annual+Meeti)
93. Xiaohua Wang, Muzhao Tian, Yuqi Zeng, Zisu Huang, Jiakang Yuan, Bowen Chen, Jingwen Xu, Mingbo Zhou, Wenhao Liu, Muling Wu, et al. Reward hacking in the era of large models: Mechanisms, emergent misalignment, challenges. arXiv preprint arXiv:2604.13602, 2026b. [find](https://scholar.google.com/scholar?q=Xiaohua+Wang%2C+Muzhao+Tian%2C+Yuqi+Zeng%2C+Zisu+Huang%2C+Jiakang+Yuan%2C+Bowen+Chen%2C+Jingwen+Xu%2C+Mingbo+Zhou%2C+Wenhao+Liu%2C+Muling+Wu%2C+et+al%2E+Reward+hacking+in+the+era+of+large+models%3A+Mechanisms%2C+emergent+misal)
94. Yiming Wang, Da Yin, Yuedong Cui, Ruichen Zheng, Zhiqian Li, Zongyu Lin, Di Wu, Xueqing Wu, Chenchen Ye, Yu Zhou, et al. Llms as scalable, general-purpose simulators for evolving digital agent training. arXiv preprint arXiv:2510.14969, 2025a. — [arXiv:2510.14969](https://arxiv.org/abs/2510.14969) [find](https://scholar.google.com/scholar?q=Yiming+Wang%2C+Da+Yin%2C+Yuedong+Cui%2C+Ruichen+Zheng%2C+Zhiqian+Li%2C+Zongyu+Lin%2C+Di+Wu%2C+Xueqing+Wu%2C+Chenchen+Ye%2C+Yu+Zhou%2C+et+al%2E+Llms+as+scalable%2C+general%2Dpurpose+simulators+for+evolving+digital+agent+trainin)
95. Yinjie Wang, Tianbao Xie, Ke Shen, Mengdi Wang, and Ling Yang. Rlanything: Forge environment, policy, and reward model in completely dynamic rl system. arXiv preprint arXiv:2602.02488, 2026c. — [arXiv:2602.02488](https://arxiv.org/abs/2602.02488) [find](https://scholar.google.com/scholar?q=Yinjie+Wang%2C+Tianbao+Xie%2C+Ke+Shen%2C+Mengdi+Wang%2C+and+Ling+Yang%2E+Rlanything%3A+Forge+environment%2C+policy%2C+and+reward+model+in+completely+dynamic+rl+system%2E+arXiv+preprint+arXiv%3A2602%2E02488%2C+2026c%2E+%E2%80%94+%5BarXiv)
96. Zhaoyang Wang, Canwen Xu, Boyi Liu, Yite Wang, Siwei Han, Zhewei Yao, Huaxiu Yao, and Yuxiong He. Agent world model: Infinity synthetic environments for agentic reinforcement learning. arXiv preprint arXiv:2602.10090, 2026d. — [arXiv:2602.10090](https://arxiv.org/abs/2602.10090) [find](https://scholar.google.com/scholar?q=Zhaoyang+Wang%2C+Canwen+Xu%2C+Boyi+Liu%2C+Yite+Wang%2C+Siwei+Han%2C+Zhewei+Yao%2C+Huaxiu+Yao%2C+and+Yuxiong+He%2E+Agent+world+model%3A+Infinity+synthetic+environments+for+agentic+reinforcement+learning%2E+arXiv+preprint+)
97. Zihan Wang, Kangrui Wang, Qineng Wang, Pingyue Zhang, Linjie Li, Zhengyuan Yang, Xing Jin, Kefan Yu, Minh Nhat Nguyen, Licheng Liu, et al. Ragen: Understanding self-evolution in llm agents via multi-turn reinforcement learning. arXiv preprint arXiv:2504.20073, 2025b. [find](https://scholar.google.com/scholar?q=Zihan+Wang%2C+Kangrui+Wang%2C+Qineng+Wang%2C+Pingyue+Zhang%2C+Linjie+Li%2C+Zhengyuan+Yang%2C+Xing+Jin%2C+Kefan+Yu%2C+Minh+Nhat+Nguyen%2C+Licheng+Liu%2C+et+al%2E+Ragen%3A+Understanding+self%2Devolution+in+llm+agents+via+multi%2Dt)
98. Ryan Wong, Jiawei Wang, Junjie Zhao, Li Chen, Yan Gao, Long Zhang, Xuan Zhou, Zuo Wang, Kai Xiang, Ge Zhang, et al. [Widesearch: Benchmarking agentic broad info-seeking](https://md.research.elamur.ai/free/en/papers/wong-2025-widesearch-benchmarking-agentic-broad-info/). arXiv preprint arXiv:2508.07999, 2025.
99. World Labs team. Marble: A multimodal world model. World Labs Technical Post, 2025. URL [https://www.worldlabs.ai/blog/marble-world-model](https://www.worldlabs.ai/blog/marble-world-model). [↗](https://www.worldlabs.ai/blog/marble-world-model)
100. Jialong Wu, Shaofeng Yin, Ningya Feng, and Mingsheng Long. Rlvr-world: Training world models with reinforcement learning. Advances in Neural Information Processing Systems, 38:125312–125350, 2026a. [find](https://scholar.google.com/scholar?q=Jialong+Wu%2C+Shaofeng+Yin%2C+Ningya+Feng%2C+and+Mingsheng+Long%2E+Rlvr%2Dworld%3A+Training+world+models+with+reinforcement+learning%2E+Advances+in+Neural+Information+Processing+Systems%2C+38%3A125312%E2%80%93125350%2C+2026a%2E)
101. Yifan Wu, Yiran Peng, Yiyu Chen, Jianhao Ruan, Zijie Zhuang, Cheng Yang, Jiayi Zhang, Man Chen, Yenchi Tseng, Zhaoyang Yu, et al. Autowebworld: Synthesizing infinite verifiable web environments via finite state machines. arXiv preprint arXiv:2602.14296, 2026b. — [arXiv:2602.14296](https://arxiv.org/abs/2602.14296) [find](https://scholar.google.com/scholar?q=Yifan+Wu%2C+Yiran+Peng%2C+Yiyu+Chen%2C+Jianhao+Ruan%2C+Zijie+Zhuang%2C+Cheng+Yang%2C+Jiayi+Zhang%2C+Man+Chen%2C+Yenchi+Tseng%2C+Zhaoyang+Yu%2C+et+al%2E+Autowebworld%3A+Synthesizing+infinite+verifiable+web+environments+via+fi)
102. Zijian Wu, Xiangyan Liu, Xinyuan Zhang, Lingjun Chen, Fanqing Meng, Lingxiao Du, Yiran Zhao, Fanshi Zhang, Yaoqi Ye, Jiawei Wang, et al. Mcpmark: A benchmark for stress-testing realistic and comprehensive mcp use. arXiv preprint arXiv:2509.24002, 2025. — [arXiv:2509.24002](https://arxiv.org/abs/2509.24002) [find](https://scholar.google.com/scholar?q=Zijian+Wu%2C+Xiangyan+Liu%2C+Xinyuan+Zhang%2C+Lingjun+Chen%2C+Fanqing+Meng%2C+Lingxiao+Du%2C+Yiran+Zhao%2C+Fanshi+Zhang%2C+Yaoqi+Ye%2C+Jiawei+Wang%2C+et+al%2E+Mcpmark%3A+A+benchmark+for+stress%2Dtesting+realistic+and+comprehen)
103. Jiannan Xiang, Yi Gu, Zihan Liu, Zeyu Feng, Qiyue Gao, Yiyan Hu, Benhao Huang, Guangyi Liu, Yichi Yang, Kun Zhou, et al. Pan: A world model for general, interactable, and long-horizon world simulation. arXiv preprint arXiv:2511.09057, 2025a. — [arXiv:2511.09057](https://arxiv.org/abs/2511.09057) [find](https://scholar.google.com/scholar?q=Jiannan+Xiang%2C+Yi+Gu%2C+Zihan+Liu%2C+Zeyu+Feng%2C+Qiyue+Gao%2C+Yiyan+Hu%2C+Benhao+Huang%2C+Guangyi+Liu%2C+Yichi+Yang%2C+Kun+Zhou%2C+et+al%2E+Pan%3A+A+world+model+for+general%2C+interactable%2C+and+long%2Dhorizon+world+simulation)
104. Jiannan Xiang, Yun Zhu, Lei Shu, Maria Wang, Lijun Yu, Gabriel Barcik, James Lyon, Srinivas Sunkara, and Jindong Chen. Uisim: An interactive image-based ui simulator for dynamic mobile environments. arXiv preprint arXiv:2509.21733, 2025b. — [arXiv:2509.21733](https://arxiv.org/abs/2509.21733) [find](https://scholar.google.com/scholar?q=Jiannan+Xiang%2C+Yun+Zhu%2C+Lei+Shu%2C+Maria+Wang%2C+Lijun+Yu%2C+Gabriel+Barcik%2C+James+Lyon%2C+Srinivas+Sunkara%2C+and+Jindong+Chen%2E+Uisim%3A+An+interactive+image%2Dbased+ui+simulator+for+dynamic+mobile+environments%2E+a)
105. Zikai Xiao, Jianhong Tu, Chuhang Zou, Yuxin Zuo, Zhi Li, Peng Wang, Bowen Yu, Fei Huang, Junyang Lin, and Zuozhu Liu. Webworld: A large-scale world model for web agent training. arXiv preprint arXiv:2602.14721, 2026. — [arXiv:2602.14721](https://arxiv.org/abs/2602.14721) [find](https://scholar.google.com/scholar?q=Zikai+Xiao%2C+Jianhong+Tu%2C+Chuhang+Zou%2C+Yuxin+Zuo%2C+Zhi+Li%2C+Peng+Wang%2C+Bowen+Yu%2C+Fei+Huang%2C+Junyang+Lin%2C+and+Zuozhu+Liu%2E+Webworld%3A+A+large%2Dscale+world+model+for+web+agent+training%2E+arXiv+preprint+arXiv%3A2)
106. Tianbao Xie, Danyang Zhang, Jixuan Chen, Xiaochuan Li, Siheng Zhao, Ruisheng Cao, Toh J Hua, Zhoujun Cheng, Dongchan Shin, Fangyu Lei, et al. [Osworld: Benchmarking multimodal agents for open-ended tasks in real computer environments](https://md.research.elamur.ai/free/en/papers/xie-2024-osworld-benchmarking-multimodal-agents-open/). Advances in Neural Information Processing Systems, 37: 52040–52094…
107. Caijun Xu, Changyi Xiao, Zhongyuan Peng, Xinrun Wang, and Yixin Cao. [Scaler: Synthetic scalable adaptive learning environment for reasoning](https://md.research.elamur.ai/free/en/papers/xu-2026-scaler-synthetic-scalable-adaptive-learning/). arXiv preprint arXiv:2601.04809, 2026a.
108. Siyuan Xu, Shiyang Li, Xin Liu, Tianyi Liu, Yixiao Li, Zhan Shi, Zixuan Zhang, Zilong Wang, Qingyu Yin, Jianshu Chen, et al. Controllable and verifiable tool-use data synthesis for agentic reinforcement learning. arXiv preprint arXiv:2604.09813, 2026b. — [arXiv:2604.09813](https://arxiv.org/abs/2604.09813) [find](https://scholar.google.com/scholar?q=Siyuan+Xu%2C+Shiyang+Li%2C+Xin+Liu%2C+Tianyi+Liu%2C+Yixiao+Li%2C+Zhan+Shi%2C+Zixuan+Zhang%2C+Zilong+Wang%2C+Qingyu+Yin%2C+Jianshu+Chen%2C+et+al%2E+Controllable+and+verifiable+tool%2Duse+data+synthesis+for+agentic+reinforceme)
109. Taofeng Xue, Chong Peng, Mianqiu Huang, Linsen Guo, Tiancheng Han, Haozhe Wang, Jianing Wang, Xiaocheng Zhang, Xin Yang, Dengchang Zhao, et al. Evocua: Evolving computer use agents via learning from scalable synthetic experience. arXiv preprint arXiv:2601.15876, 2026. — [arXiv:2601.15876](https://arxiv.org/abs/2601.15876) [find](https://scholar.google.com/scholar?q=Taofeng+Xue%2C+Chong+Peng%2C+Mianqiu+Huang%2C+Linsen+Guo%2C+Tiancheng+Han%2C+Haozhe+Wang%2C+Jianing+Wang%2C+Xiaocheng+Zhang%2C+Xin+Yang%2C+Dengchang+Zhao%2C+et+al%2E+Evocua%3A+Evolving+computer+use+agents+via+learning+from+s)
110. Sherry Yang. World models as an intermediary between agents and the real world. arXiv preprint arXiv:2602.00785, 2026. — [arXiv:2602.00785](https://arxiv.org/abs/2602.00785) [find](https://scholar.google.com/scholar?q=Sherry+Yang%2E+World+models+as+an+intermediary+between+agents+and+the+real+world%2E+arXiv+preprint+arXiv%3A2602%2E00785%2C+2026%2E+%E2%80%94+%5BarXiv%3A2602%2E00785%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2602%2E00785%29)
111. Sherry Yang, Yilun Du, Kamyar Ghasemipour, Jonathan Tompson, Leslie Kaelbling, Dale Schuurmans, and Pieter Abbeel. Learning interactive real-world simulators. arXiv preprint arXiv:2310.06114, 2023. [find](https://scholar.google.com/scholar?q=Sherry+Yang%2C+Yilun+Du%2C+Kamyar+Ghasemipour%2C+Jonathan+Tompson%2C+Leslie+Kaelbling%2C+Dale+Schuurmans%2C+and+Pieter+Abbeel%2E+Learning+interactive+real%2Dworld+simulators%2E+arXiv+preprint+arXiv%3A2310%2E06114%2C+2023%2E)
112. Bowen Ye, Rang Li, Qibin Yang, Yuanxin Liu, Linli Yao, Hanglong Lv, Zhihui Xie, Chenxin An, Lei Li, Lingpeng Kong, et al. Claw-eval: Toward trustworthy evaluation of autonomous agents. arXiv preprint arXiv:2604.06132, 2026a. — [arXiv:2604.06132](https://arxiv.org/abs/2604.06132) [find](https://scholar.google.com/scholar?q=Bowen+Ye%2C+Rang+Li%2C+Qibin+Yang%2C+Yuanxin+Liu%2C+Linli+Yao%2C+Hanglong+Lv%2C+Zhihui+Xie%2C+Chenxin+An%2C+Lei+Li%2C+Lingpeng+Kong%2C+et+al%2E+Claw%2Deval%3A+Toward+trustworthy+evaluation+of+autonomous+agents%2E+arXiv+preprint+)
113. Seonghyeon Ye, Yunhao Ge, Kaiyuan Zheng, Shenyuan Gao, Sihyun Yu, George Kurian, Suneel Indupuru, You Liang Tan, Chuning Zhu, Jiannan Xiang, et al. World action models are zero-shot policies. arXiv preprint arXiv:2602.15922, 2026b. [find](https://scholar.google.com/scholar?q=Seonghyeon+Ye%2C+Yunhao+Ge%2C+Kaiyuan+Zheng%2C+Shenyuan+Gao%2C+Sihyun+Yu%2C+George+Kurian%2C+Suneel+Indupuru%2C+You+Liang+Tan%2C+Chuning+Zhu%2C+Jiannan+Xiang%2C+et+al%2E+World+action+models+are+zero%2Dshot+policies%2E+arXiv+pr)
114. Xiao Yu, Baolin Peng, Ruize Xu, Yelong Shen, Pengcheng He, Suman Nath, Nikhil Singh, Jiangfeng Gao, and Zhou Yu. Reinforcement world model learning for llm-based agents. arXiv preprint arXiv:2602.05842, 2026. — [arXiv:2602.05842](https://arxiv.org/abs/2602.05842) [find](https://scholar.google.com/scholar?q=Xiao+Yu%2C+Baolin+Peng%2C+Ruize+Xu%2C+Yelong+Shen%2C+Pengcheng+He%2C+Suman+Nath%2C+Nikhil+Singh%2C+Jiangfeng+Gao%2C+and+Zhou+Yu%2E+Reinforcement+world+model+learning+for+llm%2Dbased+agents%2E+arXiv+preprint+arXiv%3A2602%2E0584)
115. Aohan Zeng, Xin Lv, Zhenyu Hou, Zhengxiao Du, Qinkai Zheng, Bin Chen, Da Yin, Chendi Ge, Chenghua Huang, Chengxing Xie, et al. [Glm-5: from vibe coding to agentic engineering](https://md.research.elamur.ai/free/en/papers/team-2026-5-glm-5-vibe-coding/). arXiv preprint arXiv:2602.15763, 2026.
116. Kai Zhang, Xiangchao Chen, Bo Liu, Tianci Xue, Zeyi Liao, Zhihan Liu, Xiyao Wang, Yuting Ning, Zhaorun Chen, Xiaohan Fu, et al. Agent learning via early experience. arXiv preprint arXiv:2510.08558, 2025. [find](https://scholar.google.com/scholar?q=Kai+Zhang%2C+Xiangchao+Chen%2C+Bo+Liu%2C+Tianci+Xue%2C+Zeyi+Liao%2C+Zhihan+Liu%2C+Xiyao+Wang%2C+Yuting+Ning%2C+Zhaorun+Chen%2C+Xiaohan+Fu%2C+et+al%2E+Agent+learning+via+early+experience%2E+arXiv+preprint+arXiv%3A2510%2E08558%2C+20)
117. Ziyun Zhang, Zezhou Wang, Xiaoyi Zhang, Zongyu Guo, Jiahao Li, Bin Li, and Yan Lu. [Infiniteweb: Scalable web environment synthesis for gui agent training](https://md.research.elamur.ai/free/en/papers/zhang-2026-infiniteweb-scalable-web-environment-synthesis/). arXiv preprint arXiv:2601.04126, 2026.
118. Jiale Zhao, Guoxin Chen, Fanzhe Meng, Minghao Li, Jie Chen, Hui Xu, Yongshuai Sun, Wayne Xin Zhao, Ruihua Song, Yuan Zhang, et al. Immersion in the github universe: Scaling coding agents to mastery. arXiv preprint arXiv:2602.09892, 2026. — [arXiv:2602.09892](https://arxiv.org/abs/2602.09892) [find](https://scholar.google.com/scholar?q=Jiale+Zhao%2C+Guoxin+Chen%2C+Fanzhe+Meng%2C+Minghao+Li%2C+Jie+Chen%2C+Hui+Xu%2C+Yongshuai+Sun%2C+Wayne+Xin+Zhao%2C+Ruihua+Song%2C+Yuan+Zhang%2C+et+al%2E+Immersion+in+the+github+universe%3A+Scaling+coding+agents+to+mastery%2E+a)
119. Chujie Zheng, Shixuan Liu, Mingze Li, Xiong-Hui Chen, Bowen Yu, Chang Gao, Kai Dang, Yuqiong Liu, Rui Men, An Yang, et al. [Group sequence policy optimization](https://md.research.elamur.ai/free/en/papers/zheng-2025-group-sequence-policy-optimization/). arXiv preprint arXiv:2507.18071, 2025.
120. Yuhao Zheng, Li’an Zhong, Yi Wang, Rui Dai, Kaikui Liu, Xiangxiang Chu, Linyuan Lv, Philip Torr, and Kevin Qinghong Lin. Code2world: A gui world model via renderable code generation. arXiv preprint arXiv:2602.09856, 2026. — [arXiv:2602.09856](https://arxiv.org/abs/2602.09856) [find](https://scholar.google.com/scholar?q=Yuhao+Zheng%2C+Li%E2%80%99an+Zhong%2C+Yi+Wang%2C+Rui+Dai%2C+Kaikui+Liu%2C+Xiangxiang+Chu%2C+Linyuan+Lv%2C+Philip+Torr%2C+and+Kevin+Qinghong+Lin%2E+Code2world%3A+A+gui+world+model+via+renderable+code+generation%2E+arXiv+preprint+ar)
121. Yuhang Zhou, Lizhu Zhang, Yifan Wu, Jiayi Liu, Xiangjun Fan, Zhuokai Zhao, and Hong Yan. Synthetic sandbox for training machine learning engineering agents. arXiv preprint arXiv:2604.04872, 2026. — [arXiv:2604.04872](https://arxiv.org/abs/2604.04872) [find](https://scholar.google.com/scholar?q=Yuhang+Zhou%2C+Lizhu+Zhang%2C+Yifan+Wu%2C+Jiayi+Liu%2C+Xiangjun+Fan%2C+Zhuokai+Zhao%2C+and+Hong+Yan%2E+Synthetic+sandbox+for+training+machine+learning+engineering+agents%2E+arXiv+preprint+arXiv%3A2604%2E04872%2C+2026%2E+%E2%80%94+%5Ba)
122. Kaijie Zhu, Yuzhou Nie, Yijiang Li, Yiming Huang, Jialian Wu, Jiang Liu, Ximeng Sun, Zhenfei Yin, Lun Wang, Zicheng Liu, et al. Termigen: High-fidelity environment and robust trajectory synthesis for terminal agents. arXiv preprint arXiv:2602.07274, 2026. — [arXiv:2602.07274](https://arxiv.org/abs/2602.07274) [find](https://scholar.google.com/scholar?q=Kaijie+Zhu%2C+Yuzhou+Nie%2C+Yijiang+Li%2C+Yiming+Huang%2C+Jialian+Wu%2C+Jiang+Liu%2C+Ximeng+Sun%2C+Zhenfei+Yin%2C+Lun+Wang%2C+Zicheng+Liu%2C+et+al%2E+Termigen%3A+High%2Dfidelity+environment+and+robust+trajectory+synthesis+for+)
123. Siyu Zhu, Yanbin Jiang, Hejian Sang, Shao Tang, Qingquan Song, Biao He, Rohit Jain, Zhipeng Wang, and Alborz Geramifard. Planner-r1: Reward shaping enables efficient agentic rl with smaller llms. arXiv preprint arXiv:2509.25779, 2025. — [arXiv:2509.25779](https://arxiv.org/abs/2509.25779) [find](https://scholar.google.com/scholar?q=Siyu+Zhu%2C+Yanbin+Jiang%2C+Hejian+Sang%2C+Shao+Tang%2C+Qingquan+Song%2C+Biao+He%2C+Rohit+Jain%2C+Zhipeng+Wang%2C+and+Alborz+Geramifard%2E+Planner%2Dr1%3A+Reward+shaping+enables+efficient+agentic+rl+with+smaller+llms%2E+arXi)

*Updated 31 October 2026*