Multiple Choice

An organization is developing a large language model to summarize news articles from various global sources for a diverse, international audience. Their primary ethical concern is that the model might unintentionally amplify stereotypes or misrepresent viewpoints from specific demographic or geopolitical groups. Which of the following evaluation strategies would be the most effective for identifying and quantifying this specific type of representational bias in the model's summaries?

0

1

Updated 2025-10-02

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

Evaluation in Bloom's Taxonomy

Cognitive Psychology

Psychology

Social Science

Empirical Science

Science