ReconVLA: Reconstructive Vision-Language-Action Model as Effective Robot Perceiver Paper • 2508.10333 • Published Aug 14 • 1
