QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: njuxyh_1 on February 16, 2024, 14:18

Title: MTP no active learning is suitable for MD simulation at 300K~600K?
Post by: njuxyh_1 on February 16, 2024, 14:18

Hi all.
I just wanted to make sure something. I'm a bit confused about the MTP after the initial crystalTrainingRandomDisplacement. no active learning is involved. I was wondering if it's safe to use such an MTP for molecular dynamics simulations, The temperature of the simulation is between 300k and 600k, and the crystals are still in crystal form. Could you help me out with this? Thanks!

Title: Re: MTP no active learning is suitable for MD simulation at 300K~600K?
Post by: njuxyh_1 on February 27, 2024, 05:36

Dear staff:
Please give some information or hints.

Thank you.

Title: Re: MTP no active learning is suitable for MD simulation at 300K~600K?
Post by: Julian Schneider on February 28, 2024, 22:21

It is our experience that the MTP trained on crystal displacements alone is sufficient, and can describe crystals in MD at moderate temperatures very well without active learning in almost all cases.

If you are unsure, you could run some MTP simulations of small systems, check that the simulation runs stable, and compare the forces from MTP with forces from DFT on a few snapshots from this simulations. The error should be similar to the training / testing error of the MTP.