Google

An illustrative example of generating a long text from a long context is the task of translating an entire document from one language to another, such as from Chinese to English. In this scenario, the source document represents a long input sequence (x), and the resulting translated document is a long output sequence (y).

Example of Long Text Generation Based on Long Context: Document Translation

A company wants to use an AI model to translate its entire 300-page technical manual from Chinese into English. The manual contains complex diagrams, cross-references between chapters, and a consistent set of terminology that must be maintained throughout the translated document. Analyze this scenario and explain why this translation task is fundamentally a problem of generating a long text based on a long context.

Choosing the Right AI Task

A machine learning model is tasked with translating a 100-page novel from Japanese to Spanish. Explain why this task is a clear example of generating a long text from a long context. In your explanation, explicitly identify what constitutes the long input sequence and the long output sequence in this scenario.

Analyzing a Translation Task

A developer is building a system that takes a dense, 20-page academic paper as input and rewrites it into an equally long, detailed article for a non-expert audience. The system must preserve all the original information and structure, simply rephrasing it for clarity. This task is a prime example of which type of sequence modeling problem?

Learn Before

Related