Short Answer

Deconstructing Candidate Sequences

At a certain step in text generation, a model produces the following three new candidate sequences: [The, cat, sat, on], [The, cat, sat, by], and [The, cat, sat, near]. Based on these outputs, identify two things: 1) The common prefix sequence that was extended, and 2) The set of the most probable next tokens that were used to create these new sequences.

0

1

Updated 2025-10-03

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

Analysis in Bloom's Taxonomy

Cognitive Psychology

Psychology

Social Science

Empirical Science

Science