Skip to content

Little-Podi/AdaWorld

Repository files navigation

Learning Adaptable World Models with Latent Actions

Arxiv HF Models PyTorch Python

🎬 [Project Page], 📜 [Technical Report], 🤗 [Model Weights]

Shenyuan Gao, Siyuan Zhou, Yilun Du, Jun Zhang, Chuang Gan


TL;DR: AdaWorld is a highly adaptable world model pretrained with continuous latent actions from thousands of environments, enabling efficient action transfer, adaptation, and planning with minimal finetuning.

  • Action transfer (source video → target scene)

  • Visual planning (action-agnostic vs. AdaWorld)


We introduce latent actions as a unified condition for action-aware pretraining from videos. AdaWorld can readily transfer actions across contexts without training. By initializing the control interface with the corresponding latent actions, AdaWorld can also be adapted into specialized world models efficiently and achieve significantly better planning results.

🕹️ Getting Started

❤️ Acknowledgement

Our idea is implemented based on Vista and Jafar. Thanks for their great open-source work!

⭐ Citation

If any parts of our paper and code help your research, please consider citing us and giving a star to our repository.

@article{gao2025adaworld,
  title={AdaWorld: Learning Adaptable World Models with Latent Actions}, 
  author={Gao, Shenyuan and Zhou, Siyuan and Du, Yilun and Zhang, Jun and Gan, Chuang},
  journal={arXiv preprint arXiv:2503.18938},
  year={2025}
}

📢 Contact

If you have any questions or comments, feel free to contact me through email (sygao@connect.ust.hk). Suggestions and collaborations are also highly welcome!

About

[ICML'25] The PyTorch implementation of paper: "AdaWorld: Learning Adaptable World Models with Latent Actions".

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published