Multiple Choice

A startup is building an LLM to automatically grade high school history essays. To ensure scalability and rapid deployment, they plan to align the model exclusively using AI-generated feedback. The AI feedback system will be trained to check for factual accuracy against a knowledge base, grammatical correctness, and essay length. What is the most significant risk of this alignment strategy?

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

Evaluation in Bloom's Taxonomy

Cognitive Psychology

Psychology

Social Science

Empirical Science

Science