GenesisBench: can language intelligence create physical intelligence

1. Xiao, W., et al. ENPIRE: Agentic Robot Policy Self-Improvement in the Real World. arXiv:2606.19980, Jun. 2026. arxiv.org/abs/2606.19980
2. Li, R., et al. RoboClaw: An Agentic Framework for Scalable Long-Horizon Robotic Tasks. arXiv:2603.11558, Mar. 2026. arxiv.org/abs/2603.11558