Learn Before
Case Study

Improving LLM Output Consistency

A data scientist is using a large language model to summarize long financial reports. The goal is to consistently generate a three-sentence summary highlighting the report's main finding, key risk, and future outlook. The current instruction given to the model is simply: 'Summarize the following report in three sentences, covering the main finding, key risk, and future outlook.' However, the model's outputs are unreliable; it often produces summaries of varying lengths and doesn't always structure them as requested. Based on the principle of improving model performance by providing examples directly in the input, critique the data scientist's current method and describe how you would change the instruction to achieve more consistent and correctly structured summaries.

0

1

Updated 2025-10-03

Contributors are:

Who are from:

Tags

Ch.1 Pre-training - Foundations of Large Language Models

Foundations of Large Language Models

Foundations of Large Language Models Course

Computing Sciences

Ch.3 Prompting - Foundations of Large Language Models

Ch.2 Generative Models - Foundations of Large Language Models

Evaluation in Bloom's Taxonomy

Cognitive Psychology

Psychology

Social Science

Empirical Science

Science