Pooling layers are commonly found between convolutional layers to downsample the feature representation.