Google

A zero-shot prompt provides a task instruction to a model along with a new input, without any prior examples. For instance, to classify the sentiment of a positive statement about food, the prompt could be structured as follows: 'Assume that the polarity of a text is a label chosen from {positive, negative, neutral}. Identify the polarity of the input. Input: I love the food here. It's amazing! Polarity: '

Example of a Zero-Shot Prompt for Polarity Classification (Positive Sentiment on Food)

You are tasked with using a large language model to categorize customer support tickets. Write a complete, self-contained prompt that instructs the model to classify the following customer ticket into one of these categories: 'Billing Inquiry', 'Technical Issue', or 'Account Management'. Your prompt must not include any pre-existing examples of classified tickets.

Customer Ticket Input: 'I can't seem to log into my account. I've tried resetting my password, but it's not working.'

Constructing a Zero-Shot Prompt for Ticket Classification

A marketing analyst is using a large language model to sort customer reviews for a new smartphone. They need to classify each review as 'Positive', 'Negative', or 'Neutral' without providing the model with any pre-labeled examples. The analyst's current prompt is failing to produce consistent results. Analyze the provided prompt and case details, explain why the prompt is ineffective, and then construct a revised prompt that would be more successful for this task.

Evaluating and Improving a Prompt for Sentiment Analysis

A software developer needs to categorize user feedback as either a 'bug report' or a 'feature request' using a language model. The developer has no pre-labeled examples to show the model. Which of the following prompts is the most effective and appropriate way to instruct the model to perform this task?

Learn Before

Related