Learn Before
Multiple Choice

A developer is building a system to summarize news articles. A key requirement is that the summaries must be purely objective and avoid any subjective or opinionated language. To enforce this, the developer decides to implement a penalty function. Which of the following describes a penalty that operates by evaluating the final 'surface form' of the generated text, rather than the model's internal process?

0

1

Updated 2025-10-08

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