← Back
robofoundry2026

robofoundry2026/RoboFoundry

View on GitHub ↗
Stars
30
Forks
1
Watchers
30
Open issues
1
Contributors
1
Language
—
License
—
Default branch
main
Created Sep 27, 2026Updated Sep 29, 2026

Star growth

Today—
This week—
This month—

Star history will appear here once this repo has been tracked for a couple of days.

README

RoboFoundry

System-as-Policy Evolution for Self-Learning Embodied Agents

Jingsong Liang1* · Shuhao Liao1,2* · Shizhe Zhang1 · Diyuan Hou2 · Yuxin Cai1 · Xinjian Deng3 · Chengyang He3 · Wenhui Huang1 · Runjia Tan1 · Zhidong Wang1 · Lan Yu4 · Xuesong Tian4 · Guillaume Sartoretti3 · Jie Luo2 · Yao Mu5 · Wenjun Wu2‡ · Wanhua Li1‡ · Chen Lv1‡

1Nanyang Technological University · 2Beihang University
3National University of Singapore · 4Cloud Butterfly Technology
5Shanghai Jiao Tong University

*Equal Contribution · ‡Corresponding Author

Project Page · Overview Video · Paper


RoboFoundry overview: the Act–Reflect–Repair–Promote loop for System-as-Policy evolution across robot embodiments, with results on EmbodiedBench, RoboMemArena, and LIBERO-PRO.

TODO

  • Release EmbodiedBench code.
  • Release RoboMemArena code.
  • Release LIBERO-PRO code.

Citation

@misc{liang2026robofoundry,
  title={RoboFoundry: System-as-Policy Evolution for Self-Learning Embodied Agents},
  author={Jingsong Liang and Shuhao Liao and Shizhe Zhang and Diyuan Hou and Yuxin Cai and Xinjian Deng and Chengyang He and Wenhui Huang and Runjia Tan and Zhidong Wang and Lan Yu and Xuesong Tian and Guillaume Sartoretti and Jie Luo and Yao Mu and Wenjun Wu and Wanhua Li and Chen Lv},
  year={2026},
  eprint={2609.32862},
  archivePrefix={arXiv},
  primaryClass={cs.RO},
  url={https://arxiv.org/abs/2609.32862}
}