About this paper
ResWAM combines visual residual learning in frozen DINOv3 features with a lightweight dual-branch DiT trained via conditional flow matching to improve vision-language-action policy success while lowering training and inference compute.
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,…
world-action modelsvision-language-action policiesfuture predictionvisual residual learningfrozen DINOv3 spaceresidual learning targetfeature rankdual-branch DiTaction branchresidual branch
- Upvotes
- 6
- Authors
- 5
- Published
- Nov 14, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.
Trending history
ResWAM: Learning What Changes, Not What Stays, for World-Action Models was first seen on Hugging Face · Daily papers on Nov 18, 2026. Since then it has appeared in 11 checks on 1 different day. Its best position was #17 of 30, reached on Nov 18, 05:28 UTC. It dropped off the list after Nov 18, 23:28 UTC.