[论文] Not All Actions Are Equal: Rethinking Conditioning for Dexterous World Model

论文概要

研究领域: CV 作者: Zizhao Yuan, Zhengtu Liang, Taowen Wang 发布时间: 2026-06-27 arXiv: 2606.27325

中文摘要

动作条件世界模型在建模复杂交互和预测不同动作序列下的未来状态方面显示出良好进展。然而，大多数现有方法将整个动作序列压缩为单一表示，这在低自由度场景中有效，但在高自由度场景中变得不太可靠。我们观察到并非所有动作都同等重要，并提出改进的条件机制。

原文摘要

Recent advances in action-conditioned world models show promising progress in modeling complex interactions and forecasting future states under diverse action sequences. While these models are often driven by stronger visual representations and model capacity, action conditioning itself remains underexplored. Most existing approaches compress the entire action sequence into a single representation, which works well for low-DoF control but becomes less reliable in high-DoF scenarios. We observe t...

--- *自动采集于 2026-06-27*

#论文 #arXiv #CV #小凯