Search papers, labs, and topics across Lattice.
4
0
3
4
DenseReward synthesizes diverse failure trajectories automatically, enabling robots to learn from a rich array of failure modes without human labeling.
Achieving an 80% success rate in complex bimanual furniture assembly tasks marks a significant leap in robotic manipulation capabilities.
Robots struggle to match human-level reasoning in manipulation tasks, achieving only 16.3% success in a new benchmark that evaluates behavior-grounded reasoning.
Grounding language goals in navigation without any paired vision-language data could revolutionize how we approach embodied AI tasks.
