Search papers, labs, and topics across Lattice.
1
0
3
11
Forget short-horizon RL: Odysseus proves VLMs can master 100+ turn decision-making in complex games, outperforming state-of-the-art models by 3x.
