end-to-end memory networks

A model that processes input sequences in a non-sequential order, focusing on different parts of the input at different stages of processing, typically used for natural language processing tasks.