Generating output data items using template data items

A method of generating a data item, comprising:

• receiving, by one or more processors, an input data item;
• selecting, by the one or more processors, a memory address based upon the received input data item;
• retrieving, by the one or more processors from a memory storing a plurality of template data items, a template data item based upon the selected memory address;
• processing, by an encoder neural network, the received input data item and the retrieved template data item to generate a latent variable representation;
• and
• processing, by a decoder neural network, the retrieved template data item and the latent variable representation to generate an output data item.

One or more non-transitory computer-readable storage media storing instructions that when executed by one or more computers cause the one or more computers to implement a neural network system, the neural network system comprising:

• a memory storing a plurality of template data items;
• one or more processors configured to:
• select a memory address based upon a received input data item;
• and
• retrieve a template data item from the memory based upon the selected memory address;
• an encoder neural network configured to process the received input data item and the retrieved template data item to generate a latent variable representation;
• and
• a decoder neural network configured to process the retrieved template data item and the latent variable representation to generate an output data item.