journal menu
![[Figure 2]](xx5042fig2mag.jpg)
|
|
Figure 2
(a) Data generation for training. The prior bounds and ground-truth parameters are sampled using the described procedure and the reflectivity curves are simulated from the ground-truth parameters. (b) The neural network architecture. An embedding of the reflectivity curves and the prior bounds are provided as inputs to the MLP. The MLP consists of residual blocks with batch normalization (BN), nonlinear activation (chosen to be GELU) and linear layers. |
![[Figure 2]](xx5042fig2.jpg)
 | JOURNAL OF APPLIED CRYSTALLOGRAPHY |
ISSN: 1600-5767
access
