An E(3)-equivariant graph neural network trained on MAGNDATA experimental structures predicts magnetic orders using a new primitive modulated structure representation that handles commensurate and incommensurate cases uniformly.
St¨ ohr, Journal of Magnetism and Magnetic Materials 200, 470 (1999)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cond-mat.mtrl-sci 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Universal Magnetic Structure Prediction from Atomic Coordinates with Near-Experimental Accuracy
An E(3)-equivariant graph neural network trained on MAGNDATA experimental structures predicts magnetic orders using a new primitive modulated structure representation that handles commensurate and incommensurate cases uniformly.