The authors report that they did not observe advantages when the residual function F contains only a single layer; the paper's evaluated residual functions instead use two or three layers. · CiteArk
The authors report that they did not observe advantages when the residual function F contains only a single layer; the paper's evaluated residual functions instead use two or three layers.
来源：paper:PDF page 3, Section 3.2, paragraph following Equation (2)
