Case Study

Analyzing a Prompt Optimization System

A company wants to improve its customer service chatbot. They build a system where a 'Prompt Refiner' model iteratively suggests modifications to the internal instructions used by the main chatbot. For each modified instruction, the chatbot generates a response to a test query, and this response is then rated for helpfulness on a scale of 1 to 5. The 'Prompt Refiner' uses this helpfulness rating to learn how to make better modifications in the future. Based on this scenario, analyze the system by identifying the component that corresponds to each of the following core elements of a reinforcement learning framework: Agent, Action, Environment, and Reward. Justify each of your choices.

0

1

Updated 2025-10-02

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