Short Answer

Improving a Prompt for Text Classification

A developer is using a large language model to classify customer feedback as 'Positive', 'Negative', or 'Neutral'. They use the following prompt structure:

'Here is some customer feedback: [customer feedback text]. What is your analysis?'

The model often responds with a detailed paragraph explaining the sentiment instead of a single-word label. Identify the primary weakness in this prompt and explain how you would modify it to guide the model to consistently output only one of the three desired labels.

0

1

Updated 2025-10-04

Contributors are:

Who are from:

Tags

Ch.3 Prompting - 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