Why is noise in GaussianVAE2D a torch Variable?

[hsm207]

In the GaussianVAE2D class definition in the file common_net.py, there is a method named sample.

This is how its defined:

  def sample(self, x):

    mu = self.en_mu(x)

    sd = self.softplus(self.en_sigma(x))

    noise = Variable(torch.randn(mu.size(0), mu.size(1), mu.size(2), mu.size(3))).cuda(x.data.get_device())

    return mu + sd.mul(noise), mu, sd

I don't understand why noise is defined as a Variable because we do not need to differentiate the loss function with respect to it, do we?

[mingyuliutw]

This is an implementation of the reparamerterization trick described in the VAE paper. Basically, replacing the non differentiable sampling operation with an arithmetic operation by using auxiliary stochastic variables.