TY  - RPRT
TI  - Canonical and DLPNO-based G4(MP2)XK-inspired composite wavefunction methods parametrized against large and chemically diverse training sets: Are they more accurate and/or robust than double hybrid DFT?
AU  - Emmanouil Semidalas
AU  - Jan M. L. Martin
PY  - 2020
DO  - 10.1021/acs.jctc.0c00189
UR  - https://arxiv.org/abs/2006.04742
ID  - 2006.04742
ER  -