Activity (Process)

Iterative Context Update in Autoregressive Generation

In autoregressive text generation, the process is iterative. After a set of one or more tokens is predicted in a single step, these new tokens are appended to the existing context. This newly expanded context is then used as the basis for the next prediction cycle, allowing the model to continue generating the sequence.

0

1

Updated 2025-10-10

Contributors are:

Who are from:

Tags

Ch.5 Inference - Foundations of Large Language Models

Foundations of Large Language Models

Foundations of Large Language Models Course

Computing Sciences