Variational autoencoders (VAES) are generative autoencoders - they generate new data from the input. What is the coding fed into the decoder in these models? Instruction: Choose the option that best answers the question. The latent representation produced by the encoder represented in lower dimensionality The decoder directly takes in the input image specified using statistical parameters - VAES do not use encoder models The mean and standard deviation of the input distribution of the data used to train the model A random sample from a Gaussian distribution with mean and standard deviation produced by the encoder

Variational autoencoders (VAES) are generative autoencoders - they generate new data from the input. What is the coding fed into the decoder in these models? Instruction: Choose the option that best answers the question. The latent representation produced by the encoder represented in lower dimensionality The decoder directly takes in the input image specified using statistical parameters - VAES do not use encoder models The mean and standard deviation of the input distribution of the data used to train the model A random sample from a Gaussian distribution with mean and standard deviation produced by the encoder