Case Study

Prioritizing Chatbot Alignment Strategies

A company's new customer service chatbot, built on a powerful language model, exhibits two key issues. First, when given a direct command like 'Summarize my last three orders,' it often continues the sentence with plausible but irrelevant text instead of executing the command. Second, when a user expresses frustration, the model provides factually correct but unempathetic and unhelpful replies. The development team must decide which alignment approach to prioritize for each issue. Analyze this scenario and determine which of the two fundamental alignment approaches is best suited to address the first issue, and which is best for the second. Justify your choices.

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

Ch.4 Alignment - Foundations of Large Language Models

Ch.2 Generative Models - Foundations of Large Language Models

Analysis in Bloom's Taxonomy

Cognitive Psychology

Psychology

Social Science

Empirical Science

Science