Search papers, labs, and topics across Lattice.
1
0
3
Equivariant atomistic models like MACE aren't just more accurate; they learn linearly disentangled representations, making downstream tasks far more sample-efficient.
