Before the Body Moves: Learning Anticipatory Joint Intent for Language-Conditioned Humanoid Control

ArXi:2605.14417v1 Announce Type: cross Natural language is an intuitive interface for humanoid robots, yet streaming whole-body control requires control representations that are executable now and anticipatory of future physical transitions. Existing language-conditioned humanoid systems typically generate kinematic references that a low-level tracker must repair reactively, or use latent/action policies whose outputs do not explicitly encode upcoming contact changes, transfers, and balance preparation.