Search papers, labs, and topics across Lattice.
2
0
3
E(3) equivariance and derivative-aware training enables a GNN to predict XANES spectra with high fidelity, opening the door to ML-assisted spectroscopy and accelerated materials discovery.
LLMs can guide the generation of realistic crystal structures by first producing natural language descriptions of unit cell layouts, which are then translated into atomic coordinates by a flow model.
