Fig. 3From: Neutron: an attention-based neural decompilerAn example of training data augmentationBack to article page