Geometric Action Model for Robot Policy Learning Repurposes a geometric foundation model as one backbone for perception, prediction, and action. 1.…

A 1.4B-parameter robot policy model unifies perception, prediction, and action under a single geometric foundation model backbone, achieving 85.5% on LIBERO-Plus with 6.9 ms inference latency. That is 55x faster than reported baselines, making real-time deployment on physical hardware more practical.