Search papers, labs, and topics across Lattice.
1
0
3
Max-Plus networks, despite their interpretability, can be efficiently trained by exploiting the algebraic sparsity of their subgradients, leading to faster updates.
