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)

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?

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.

Learn Before

Related