A new open-source toolkit automates active learning and dataset management for neuroevolution potentials, with a CsPbI3 case study showing comparable accuracy to hand-curated NEP models.
Meanwhile, the-for–-filterparameter allows users to enable or disable filtering based on the minimum bond length (disabled by default)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
NepTrain and NepTrainKit: Automated Active Learning and Visualization Toolkit for Neuroevolution Potentials
A new open-source toolkit automates active learning and dataset management for neuroevolution potentials, with a CsPbI3 case study showing comparable accuracy to hand-curated NEP models.