Fig. 6: A comparison of the interface properties of water predicted using short-range (SR) or long-range (LR) models and without (T = 0) or with one massage passing layer (T = 1).
From: Latent Ewald summation for machine learning of long-range interactions
a shows a snapshot of the thinner water slab configuration, b shows the water density profile, and c shows average cosine (c) the average \(\cos (\theta )\) for the angle formed by the water dipole moment and z-axis. DFT results are from ref. 4. d, e and f show the snapshot, the water density, and \(\langle \cos (\theta )\rangle\) for the thicker water slab, respectively.
