Search papers, labs, and topics across Lattice.
1
0
3
4
SLPO enables latent reasoning models to leverage outcome-reward learning, dramatically enhancing their performance and efficiency in complex reasoning tasks.
