作者位次 #4/4ICLR 2026Poster· Reinforcement Learning / Batch/Offline
Cross-Embodiment Offline Reinforcement Learning for Heterogeneous Robot Datasets
Haruki Abe · Takayuki Osa · YUSUKE Mukuta · Tatsuya Harada
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Haruki Abe · Takayuki Osa · YUSUKE Mukuta · Tatsuya Harada
Naoki Shitanda · Motoki Omura · Tatsuya Harada · Takayuki Osa
Naoki Morihira · Amal Nahar · Kartik Bharadwaj · Yasuhiro Kato · Akinobu Hayashi · Tatsuya Harada
Zhongyi Zhou · Kohei Uehara · Haoyu Zhang · Jingtao Zhou · Lin Gu · Ruofei Du · Zheng Xu · Tatsuya Harada
Prior work synthesizes tool-use LLM datasets by first generating a user query, followed by complex tool-use annotations like DFS. This inherently leads to inevitable annotation failures and low efficiency in data generation. We introduce ToolGrad, an agentic framework that inverts this paradigm....
Yasuhiro Kato · Thomas Westfechtel · Jen-Yen Chang · Naoki Morihira · Akinobu Hayashi · Tatsuya Harada · Takayuki Osa
Tanachai Anakewat · Takayuki Osa · Tatsuya Harada