FIG 2 - uploaded by Ji Woong Yu
Content may be subject to copyright.
(a) The prepared molecular dynamics snapshots are randomly sampled to compose the input batch. (b) The schematic diagram of the network architecture, (c) Visualization of attention array (QK T ) in 2D fixed boundary LJ (N = 64) system. The color of the nodes denotes the difference between the true and predicted energy for which the corresponding color bar is shown at the right bottom of (c). Each head shows a different pattern of attention. Note that only a few highest attraction values starting from one node or those with an attraction value higher than 0.08 are visualized. Also, the transparency of the edge indicates the relative intensity of attention.
Source publication
The combination of neural network potential (NNP) with molecular simulations plays an important role in an efficient and thorough understanding of a molecular system's potential energy surface (PES). However, grasping the interplay between input features and their local contribution to NNP is growingly evasive due to heavy featurization. In this wo...
Similar publications
In view of the fundamental distinction between the force-controlled model and the displacement-controlled model in buckling problems of structures and the complexity of the asymptotic post-buckling analysis traditionally based on the force-controlled model, alternatively, we provide a straightforward theoretical procedure of the energy method for t...
