fig1

A deep neural network potential model for transition metal diborides

Figure 1. (A) Illustration of the DFT dataset generation process using a concurrent learning scheme in the DP-GEN software. The lower part of the image demonstrates the coverage of the configurational space by the current model, with accurate, candidate, and inaccurate regions distinguished by different colors; (B) Comparison of energy predictions between the DP model and DFT calculations; (C) Error distribution in predicted energy; (D) Comparison of force predictions between the DP model and DFT calculations; (E) Error distribution in predicted force. DFT: Density functional theory; DP-GEN: Deep Potential GENerator; DP: Deep Potential.

Journal of Materials Informatics
ISSN 2770-372X (Online)
Follow Us

Portico

All published articles are preserved here permanently:

https://www.portico.org/publishers/oae/

Portico

All published articles are preserved here permanently:

https://www.portico.org/publishers/oae/