Search papers, labs, and topics across Lattice.
1
0
Discretizing reward models can significantly enhance policy performance by reducing oversensitivity without sacrificing discriminative ability.
