Search papers, labs, and topics across Lattice.
1
10
3
4
A 7B model trained with reinforcement learning beats GPT-4o on robotic reasoning tasks involving low-level action control.
