Challenge of Prompting LLMs for Many-Category Classification
A significant difficulty in using Large Language Models for classification arises when the number of potential categories is large. In such scenarios, designing effective prompts to guide the model becomes a considerable challenge.
0
1
Tags
Ch.3 Prompting - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Related
Example of a Complete Prompt for Polarity Classification
Components of an Instruction-based Prompt
Zero-Shot Learning with LLMs
Example of a Zero-Shot Prompt for Polarity Classification (Negative Sentiment)
Examples of Instruction-based Prompts for Polarity Classification
Using Descriptive Prompts for Novel Classification Tasks
Challenge of Prompting LLMs for Many-Category Classification
Example of a Zero-Shot Prompt for Polarity Classification (Positive Sentiment)
Example of a Zero-Shot Prompt for Polarity Classification (Positive Sentiment on Food)
Adapting Prompt Detail to an LLM's Task Familiarity
A developer needs a large language model to classify incoming customer support tickets. The goal is to sort each ticket into one of three specific categories: 'Technical Issue', 'Billing Inquiry', or 'General Feedback'. Which of the following prompts is best structured to achieve this task reliably and consistently?
Diagnosing Ineffective Prompt Instructions
Crafting an Instruction for a Novel Task
Instructing LLMs with Detailed Descriptions
Proposal for LLM-based Classification System
A financial services company is developing a system to classify customer support emails into one of three categories: 'Account Inquiry', 'Technical Issue', or 'Complaint'. They are considering two different prompting strategies for their Large Language Model.
- Strategy A: "Classify the following email into one of these categories: Account Inquiry, Technical Issue, Complaint. Email: [email text]"
- Strategy B: "Analyze the following email and determine if it is an Account Inquiry, a Technical Issue, or a Complaint. If it does not fit any of these, label it as 'Other'. Email: [email text]"
Which of the following statements best evaluates the potential risks associated with these strategies?
Challenge of Prompting LLMs for Many-Category Classification
Evaluating LLM vs. Fine-Tuned Models for Classification
A team is deploying a text classification system using a large language model. They encounter several unexpected behaviors. Match each observed behavior with the most likely underlying issue.