Search papers, labs, and topics across Lattice.
3
0
5
Even top-performing voice MLLMs break down when forced to talk science, routinely failing to verbalize symbolic notation, decode technical jargon, and adapt across progressive multi-turn interactions.
A unified policy enables humanoid robots to robustly interact with complex environments, achieving record-breaking success rates in dynamic terrain and command scenarios.
Legged robots can now recover from sensor noise and crazy user commands with 10x greater reliability, thanks to a new method that respects the robot's competence boundaries.
