Search papers, labs, and topics across Lattice.
3
0
9
1
LLMs can learn to play complex games far more effectively by co-evolving a skill bank with a decision-making agent, enabling consistent long-horizon decision-making.
Generate semantically aligned, high-fidelity music for videos with unprecedented speed and control by combining autoregressive planning and diffusion.
Counterfactual human preference labels dramatically improve robot navigation safety, reducing collisions by nearly 50% in simulation and boosting real-world goal success by 24%.
