Search papers, labs, and topics across Lattice.
2
0
5
A 7B model, trained with reinforcement learning to reason about robotic manipulation, outperforms 72B-scale general MLLMs and even OpenAI's closed-source models in process supervision and failure detection.
Achieve real-time embodied manipulation with large 3D vision models using a novel asynchronous architecture that boosts success rates by up to 51.4% while simultaneously reducing inference time.
