A 0.5B parameter latent world-action model trains end-to-end on a single GPU and reaches 90.48% average success on 50 RoboTwin 2.0 tasks with a language-free Visual Transition Token for task specification.
Predictive inverse dynamics models are scalable learners for robotic manipulation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.RO 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
LiLa-WAM: Lightweight Latent Reasoning World-Action Model for Robotic Manipulation
A 0.5B parameter latent world-action model trains end-to-end on a single GPU and reaches 90.48% average success on 50 RoboTwin 2.0 tasks with a language-free Visual Transition Token for task specification.