arXiv preprint / cs.RO

A Balanced Data Diet: Addressing the Exploration Bottleneck in Mega-Scale RL for Robot Control

作者：Octi Zhang, Mateo Guaman Castro, Patrick Yin, Ignacio Dagnino, Abhishek Gupta, Rosario Scalise, Byron Boots · 单位：University of Washington · 会议/期刊：arXiv preprint · 方向：cs.RO · 发布日期：2026-10-09

2026-10-09🧠 机器人学习
新标签打开 AlphaXiv
A Balanced Data Diet: Addressing the Exploration Bottleneck in Mega-Scale RL for Robot Control figure
AlphaXiv 中文概览（可滚动查看）