Fig. 4: Conceptual DFT quantities predicted by the AIMNet-NSE model.
From: Teaching a neural network to attach and detach electrons from molecules
Correlation between DFT PBE0/ma-def2-SVP and AIMNet-NSE predictions for electronegativity (χ), chemical hardness (η) and electrophilicity index (\(\omega\)), Fukui coefficients for nucleophilic (\({f}_{a}^{+}\)), for electrophilic (\({f}_{a}^{-}\)) and radical (\({f}_{a}^{\pm }\)) attacks and three corresponding condensed philicity indexes (ωa) for Ions-16 data set.
