Skip to content

world-gymnast/world-gymnast

 
 

Repository files navigation

World-Gymnast: Training Robots with Reinforcement Learning in a World Model

World-Gymnast fine-tunes vision-language-action policies by rolling out actions in a learned world model and scoring imagined trajectories with a vision-language model. It improves real-robot performance, supports training with distractors and novel language, and enables test-time training plus iterative world-model updates.

World-Gymnast main figure

Setup

See SETUP.md for environment setup and dependencies.

Running World-Gymnast

Example training script:

  • examples/run_openvla_oft_rl_worldgym.sh

Data

Training data is stored as JSON annotations plus PNG images.

Example JSON:

{"instruction": "lift eggplant", "partial_credit_criteria": "the robot makes contact with the eggplant"}

Project Website

https://world-gymnast.github.io/

Paper

https://arxiv.org/abs/2602.02454

Citation

@misc{sharma2026worldgymnasttrainingrobotsreinforcement,
      title={World-Gymnast: Training Robots with Reinforcement Learning in a World Model},
      author={Ansh Kumar Sharma and Yixiang Sun and Ninghao Lu and Yunzhe Zhang and Jiarao Liu and Sherry Yang},
      year={2026},
      eprint={2602.02454},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2602.02454},
}

Acknowledgements

About

World-Gymnast: Training Robots with Reinforcement Learning in a World Model

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 99.5%
  • Shell 0.5%