Convolutional layers are not connected to every pixel in the…

Convolutional layers are not connected to every pixel in the input image, but only to pixels in their receptive fields. As multiple layers of convolutions are stacked, the network is able to concentrate on [____] in the first hidden layer, early on in the model process.   Fill in the blank [ ]: An example is also provided for each choice. 

Jasmine sharpens her sword against a spinning grinding wheel…

Jasmine sharpens her sword against a spinning grinding wheel, which can be modeled as a uniform cylinder. The 25 kg-wheel has a radius of 20 cm and is spinning at 200 rpm. If after 15 s of sharpening the grinding wheel is then rotating at 100 rpm, what is the torque applied to the grinding wheel by Jasmine?