CNN Receptive Field Growth Across Layers (2D)

A convolutional neural network never looks at the whole image at once inside a single layer — each output unit only "sees" a small, growing patch of the input, called its receptive field. This 2D companion visualizer draws every stacked layer's receptive-field boundary as a nested contour directly on the original 32×32 input grid, alongside a shrinking-square strip of the true forward feature-map sizes, so growth-backward and shrinkage-forward are visible in the same frame. Adjust the number of layers, kernel size, stride and dilation and watch the nested boundaries expand or contract live, with the effective receptive-field size, output feature-map resolution and cumulative stride ("jump") read out numerically.