Search papers, labs, and topics across Lattice.
3
0
5
2
Allocating extra computation during decision-making boosts robot manipulation success rates, revealing a critical strategy for handling complex tasks.
Internal dual modality prompting can significantly enhance spatial detail while preserving spectral fidelity in hyperspectral pansharpening.
Generalist robot policies can achieve 95% success rates on real-world manipulation tasks by continually learning from a fleet of robots, even in the face of distribution shifts and long-tail failures.
