AI & ML
ResWAM: Learning What Changes, Not What Stays, for World-Action Models
arXiv cs.RO0 views
arXiv:2611.19559v1 Announce Type: new Abstract: World-action models (WAMs) augment vision-language-action policies (VLAs) with future prediction so that the policy can anticipate how its actions will affect the physical world. Existing WAMs supervise complete future observations. However, observations typically change little over an action chunk, so the future loss is dominated by copying static background content rather than learning task-relevant robot and object motion, limiting policy learning. Moreover, existing WAMs rely on large video generation models or separate world models, incurring heavy training and inference compute that hinders closed-loop deployment. To improve policy performance, we propose visual residual learning, supervising only the feature change from the current frame to the future frame in a frozen DINOv3 space. Compared with complete future features, this residual learning target yields a higher feature rank with richer and more diverse representations of scene changes, and spatially concentrates on the robot arm and manipulated objects rather than the background. To further improve computational efficiency, we present ResWAM, a residual world-action model with 2.3B parameters in total, including a lightweight 0.5B dual-branch DiT that jointly trains an action branch and a residual branch via conditional flow matching. Without robot pretraining (i.e., pretraining on external robot action trajectories beyond the benchmark demonstrations), ResWAM achieves top average success rates on LIBERO-Plus (81.1%), RoboTwin 2.0 (93.8%), and RoboDojo (9.40% success, 13.41 score) at the lowest training (2.7-4.0 TFLOPs) and inference compute (1.0-1.4 TFLOPs), and leads real-robot Open, Put, and Stack (80.0% average, +10.0% over $\pi_{0.5}$) at 92.7 ms latency.
Source:https://arxiv.org/abs/2611.19559PreviousCooperative Gradient Estimation for Multi-Robot Source Seeking with Heteroskedastic MeasurementsNextTowards Probabilistically Safe Reinforcement Learning: Theory and Deployment on Humanoid and Aerial Robots
Related posts
TacBridge: Learning Unified Tactile Representations through Cross-Sensor Correspondence
0arXiv cs.RO
Transferable Neural CBFs via Generalized Closure
0arXiv cs.RO
QTOP: Scaling Humanoid Motion Tracking with Quantile Critics
0arXiv cs.RO
Non-Imaging Optical Tactile Sensors for Robotics: Transduction Principles, System Integration, and Robotic Applications
0arXiv cs.RO
Robotic Reward Models in the Foundation-Model Era
0arXiv cs.RO
Selective Sensor Acquisition for Multimodal Place Recognition
0arXiv cs.RO