第001/21页（英文原文）DreamDojo: A Generalist Robot World Model from Large-Scale Human VideosShenyuan Gao1,2† William Liang1,3† Kaiyuan Zheng1,4* Ayaan Malik1,5* Seonghyeon Ye1,6 SihyunY u 6 \mathbf { Y u } ^ { 6 }Yu
DreamDojo：基于大规模人类视频的通用机器人世界模型【预训练：流匹配损失（预测未来帧）】【在pretrain阶段之前预先训练好Latent Action Model（LAM）】
发布时间：2026/7/18 3:11:10
第001/21页（英文原文）DreamDojo: A Generalist Robot World Model from Large-Scale Human VideosShenyuan Gao1,2† William Liang1,3† Kaiyuan Zheng1,4* Ayaan Malik1,5* Seonghyeon Ye1,6 SihyunY u 6 \mathbf { Y u } ^ { 6 }Yu
