Bklight999/World-Knowledge

World Knowledge Dataset and Model

This training model and dataset are the official artifact for the paper [Training LLM Agents for Spontaneous, Reward-Free Self-Evolution via World Knowledge Exploration](https://huggingface.co/papers/2604.18131).

📄 Paper Information

• —Title: Training LLM Agents for Spontaneous, Reward-Free Self-Evolution via World Knowledge Exploration
• —Authors: Qifan Zhang, Dongyang Ma, Tianqing Fang, Jia Li, Jing Tang, Nuo Chen, Haitao Mi, Yan Wang
• —Hugging Face Paper Page: https://huggingface.co/papers/2604.18131
• —arXiv: arXiv:2604.18131

💻 Code Repository

The official codebase for the web-agent pipeline, including data generation, preprocessing, and evaluation scripts, can be found on GitHub: [Bklight999/world-knowledge](https://github.com/Bklight999/world-knowledge)

📝 Citation

If you use this repo or our code, please cite our paper:

@article{zhang2026training,
  title={Training LLM Agents for Spontaneous, Reward-Free Self-Evolution via World Knowledge Exploration},
  author={Zhang, Qifan and Ma, Dongyang and Fang, Tianqing and Li, Jia and Tang, Jing and Chen, Nuo and Mi, Haitao and Wang, Yan},
  journal={arXiv preprint arXiv:2604.18131},
  year={2026}
}