4 papers
One Ring to Rule Them All: Unifying Group-Based RL via Dynamic Power-Mean Geometry
Weisong Zhao, Tong Wang, Zichang Tan +11
Group-based reinforcement learning has evolved from the arithmetic mean of GRPO to the geometric mean of GMPO. While GMPO improves stability by constraining a conservative objectiv…
FieldGen: From Teleoperated Pre-Manipulation Trajectories to Field-Guided Data Generation
Wenhao Wang, Kehe Ye, Xinyu Zhou +9
Large-scale and diverse datasets are vital for training robust robotic manipulation policies, yet existing data collection methods struggle to balance scale, diversity, and quality…
Uncertainty-aware Reward Design Process
Yang Yang, Xiaolu Zhou, Bosong Ding +1
Designing effective reward functions is a cornerstone of reinforcement learning (RL), yet it remains a challenging process due to the inefficiencies and inconsistencies inherent in…
Genie Centurion: Accelerating Scalable Real-World Robot Training with Human Rewind-and-Refine Guidance
Wenhao Wang, Jianheng Song, Chiming Liu +13
While Vision-Language-Action (VLA) models show strong generalizability in various tasks, real-world deployment of robotic policy still requires large-scale, high-quality human expe…