Which token marks the end of a message in GPT-4?

Study for the Hugging Face Agent Certification. Prepare with interactive quizzes and multiple-choice questions, complete with explanations and hints. Ace your exam!

Multiple Choice

Which token marks the end of a message in GPT-4?

Explanation:
End-of-text is the signal that finishes a message. It marks the boundary after the content and tells the model to stop generating, effectively ending that piece of text. In many teaching contexts for GPT-4, the token string <|endoftext|> is used as this end marker, which is why it’s the best choice here. The start-of-text marker signals the opposite boundary—the beginning of content—so it wouldn’t indicate the end. Other options like end-of-message or a shorter end token aren’t the standard termination signal in this setup, so they’re not the correct choice for marking the end of a message.

End-of-text is the signal that finishes a message. It marks the boundary after the content and tells the model to stop generating, effectively ending that piece of text. In many teaching contexts for GPT-4, the token string <|endoftext|> is used as this end marker, which is why it’s the best choice here. The start-of-text marker signals the opposite boundary—the beginning of content—so it wouldn’t indicate the end. Other options like end-of-message or a shorter end token aren’t the standard termination signal in this setup, so they’re not the correct choice for marking the end of a message.

Subscribe

Get the latest from Passetra

You can unsubscribe at any time. Read our privacy policy