Can you explain this? What are the correct explanations for pooling layers? $($Multiple Answers Possible$)$ A$.$ An average pooling layer is differentiable since it is a linear operation. B$.$ A max pooling layer has strictly fewer learnable parameters than a convolutional layer. C$.$ A max pooling layer can be used between convolutional layers without breaking the computation graph. D$.$ A pooling layer must include a non$-$linear activation function.