FlashcardBeginner

Next-Token Prediction

LLMs work by repeatedly predicting the most likely next token given the preceding text.

Question

How does an LLM generate text at its core?

Click to reveal answer