arXiv Robotics — research abstracts· Junyan Li, Ruizhi Li, Yu Liu, Xiangshuo Liu, Mingchao Sun, Hongyu Pan, Mu Xu, Lue Fan, Zhaoxiang Zhang·· 10 days agoEditorial score75
DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training
DreamTrue: Action-Faithful Robot World Model with Counterfactual Post-Training
Summary
DreamTrue is a multi-view, cross-embodiment robot world model that enables action-faithful and physically plausible video prediction. It addresses calibration and interaction coverage challenges through offline geometric calibration and counterfactual post-training. The model reduces human-assessed interaction defects from 48.12% to 6.25% and ranks first in the AgiBot World Challenge 2026.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
Source：arXiv Robotics — research abstracts · arxiv.org