Fig. 4From: A machine learning correction for DFT non-covalent interactions based on the S22, S66 and X40 benchmark databasesNCIs calculated by DFT M06-2X (left) and DFT-GRNN (right) versus benchmark values. The insets are the deviation (calculated-benchmark) distribution relative to the benchmark values in each calculation (training set: red; test set: blue)Back to article page