SOURCE-LINKED INTELLIGENCE

AcrossWAM1.0:A Modular Latent World-Action Stack for Compact Robot Policies

arXiv · AI, language, vision and robotics · article · Aug 30, 2026 · UTC

Latent world-action models avoid rendering future pixels by predicting an action-relevant visual subgoal in feature space. LaWAM established this formulation, but its original presentation left the world model, multimodal backbone, and deployment checkpoint tightly coupled. We introduce AcrossWAM1.0, a modularization and scaling study of this latent world-action stack. Rather than presenting latent subgoals as a new algorithm, we make the module boundary explicit: a policy adapter produces latent-action and action-generation contexts; a retained latent world decoder grounds the predicted trans

Read original source ↗ Open in workspace

recordType

paper

region

Global