Figure 4
From: Virtual data augmentation method for reaction prediction
Visualization of attention weight before and after Hiyama reaction augmentation. The horizontal axis contains two reactants and reagents, and the vertical axis is the product. (a) SMILES:CC(=O)c1ccc(I)cc1.F[Si](c1ccccc1)(c1ccccc1)c1ccccc1.[F].[K+]>>CC(=O)c1ccc(c2ccccc2)cc1. (b) SMILES:CC[Si](Cl)(Cl)c1ccc(C)cc1.N#Cc1ccc(Br)cc1.[F-].[K+]>>Cc1ccc(-c2ccc(C#N)cc2)cc1.
