Fig. 2From: Benchmarks for interpretation of QSAR modelsInterpretation performance of models trained on the N data setBack to article page