Matches in SemOpenAlex for { <https://semopenalex.org/work/W3121325512> ?p ?o ?g. }
- W3121325512 abstract "Neural network (NN)-based interatomic potentials provide fast prediction of potential energy surfaces with the accuracy of electronic structure methods. However, NN predictions are only reliable within well-learned training domains, with unknown behavior when extrapolating. Uncertainty quantification through NN committees identify domains with low prediction confidence, but thoroughly exploring the configuration space for training NN potentials often requires slow atomistic simulations. Here, we employ adversarial attacks with a differentiable uncertainty metric to sample new molecular geometries and bootstrap NN potentials. In combination with an active learning loop, the extrapolation power of NN potentials is improved beyond the original training data with few additional samples. The framework is demonstrated on multiple examples, leading to better sampling of kinetic barriers and collective variables without extensive prior data on the relevant geometries. Adversarial attacks are new ways to simultaneously sample the phase space and bootstrap NN potentials, increasing their robustness and enabling a faster, accurate prediction of potential energy landscapes." @default.
- W3121325512 created "2021-02-01" @default.
- W3121325512 creator A5018079613 @default.
- W3121325512 creator A5043932662 @default.
- W3121325512 creator A5049028674 @default.
- W3121325512 date "2021-01-27" @default.
- W3121325512 modified "2023-09-27" @default.
- W3121325512 title "Adversarial Attacks on Uncertainty Enable Active Learning for Neural Network Potentials." @default.
- W3121325512 cites W1973858832 @default.
- W3121325512 cites W1975997599 @default.
- W3121325512 cites W1985272737 @default.
- W3121325512 cites W1988091937 @default.
- W3121325512 cites W2000957843 @default.
- W3121325512 cites W2016351523 @default.
- W3121325512 cites W2025444507 @default.
- W3121325512 cites W2029413789 @default.
- W3121325512 cites W2030687437 @default.
- W3121325512 cites W2031430816 @default.
- W3121325512 cites W2037464828 @default.
- W3121325512 cites W2044028691 @default.
- W3121325512 cites W2053117030 @default.
- W3121325512 cites W2055526416 @default.
- W3121325512 cites W2057858097 @default.
- W3121325512 cites W2058370262 @default.
- W3121325512 cites W2067718414 @default.
- W3121325512 cites W2083415705 @default.
- W3121325512 cites W2083956694 @default.
- W3121325512 cites W2086513155 @default.
- W3121325512 cites W2086957099 @default.
- W3121325512 cites W2104489082 @default.
- W3121325512 cites W2122427541 @default.
- W3121325512 cites W2130306094 @default.
- W3121325512 cites W2135293965 @default.
- W3121325512 cites W2149655632 @default.
- W3121325512 cites W2164411961 @default.
- W3121325512 cites W2335163884 @default.
- W3121325512 cites W2337496963 @default.
- W3121325512 cites W2406709014 @default.
- W3121325512 cites W2420659973 @default.
- W3121325512 cites W2470768373 @default.
- W3121325512 cites W2524276051 @default.
- W3121325512 cites W2527189750 @default.
- W3121325512 cites W2601081289 @default.
- W3121325512 cites W2605925159 @default.
- W3121325512 cites W2606142660 @default.
- W3121325512 cites W26088913 @default.
- W3121325512 cites W2620687153 @default.
- W3121325512 cites W2742127985 @default.
- W3121325512 cites W2788686132 @default.
- W3121325512 cites W2794979247 @default.
- W3121325512 cites W2804180102 @default.
- W3121325512 cites W2883583109 @default.
- W3121325512 cites W2884430236 @default.
- W3121325512 cites W2889326414 @default.
- W3121325512 cites W2892094491 @default.
- W3121325512 cites W2901155091 @default.
- W3121325512 cites W2922733622 @default.
- W3121325512 cites W2923693308 @default.
- W3121325512 cites W2949208225 @default.
- W3121325512 cites W2953641781 @default.
- W3121325512 cites W2963207607 @default.
- W3121325512 cites W2964059111 @default.
- W3121325512 cites W2964121744 @default.
- W3121325512 cites W2964153729 @default.
- W3121325512 cites W2968923792 @default.
- W3121325512 cites W2971894235 @default.
- W3121325512 cites W2972246420 @default.
- W3121325512 cites W2974593196 @default.
- W3121325512 cites W2976720228 @default.
- W3121325512 cites W2996443485 @default.
- W3121325512 cites W3000508506 @default.
- W3121325512 cites W3004981459 @default.
- W3121325512 cites W3019725847 @default.
- W3121325512 cites W3035559885 @default.
- W3121325512 cites W3036490780 @default.
- W3121325512 cites W3084103887 @default.
- W3121325512 cites W3090555547 @default.
- W3121325512 cites W3119329252 @default.
- W3121325512 cites W3119697748 @default.
- W3121325512 cites W1993387039 @default.
- W3121325512 hasPublicationYear "2021" @default.
- W3121325512 type Work @default.
- W3121325512 sameAs 3121325512 @default.
- W3121325512 citedByCount "0" @default.
- W3121325512 crossrefType "posted-content" @default.
- W3121325512 hasAuthorship W3121325512A5018079613 @default.
- W3121325512 hasAuthorship W3121325512A5043932662 @default.
- W3121325512 hasAuthorship W3121325512A5049028674 @default.
- W3121325512 hasConcept C104317684 @default.
- W3121325512 hasConcept C105795698 @default.
- W3121325512 hasConcept C119857082 @default.
- W3121325512 hasConcept C127413603 @default.
- W3121325512 hasConcept C132459708 @default.
- W3121325512 hasConcept C154945302 @default.
- W3121325512 hasConcept C176217482 @default.
- W3121325512 hasConcept C185592680 @default.
- W3121325512 hasConcept C21547014 @default.
- W3121325512 hasConcept C32230216 @default.
- W3121325512 hasConcept C33923547 @default.
- W3121325512 hasConcept C37736160 @default.