Skip to main content
Practice

Prompt Engineering Wrap-up

Let's summarize the key points of prompt engineering we've learned so far.


Components of a Prompt​

Instruction​

The part that specifies the task you want the AI to perform

Context​

The background information and situation to refer to when performing a task based on the instruction

Input Data​

The data the AI analyzes or references while performing tasks

Output Directives​

The format of the results required in the prompt


Prompt Engineering​

The technique of writing questions or instructions clearly and specifically to obtain the desired response from the AI


Hallucination​

The phenomenon where AI generates incorrect information that doesn’t exist in reality


Delimiters​

Characters or strings (multiple characters) like ###, --- that clearly separate and structure information within a prompt


Prompting Techniques​

Zero-shot Prompting​

Handling new tasks that the AI has not specifically trained on, based on pre-trained knowledge


Few-shot Prompting​

Providing AI with a few examples and requesting it to perform a specific task based on them


Chain of Thought Prompting​

Assisting AI to solve complex problems in a step-by-step manner and encouraging it to explain the reasoning process


Prompt Chaining​

Using the output from one prompt as the input to the next prompt


Least-to-Most Prompting​

Gradually expanding prompts from simple questions to complex ones


Analogical Prompting​

Referencing similar situations or cases


Knowledge Generation​

Generating useful information concerning the question before creating the final response


Role Prompting​

Guiding AI to generate answers based on a specific role


Self-consistency Prompting​

Introducing or generating multiple lines of reasoning in advance and encouraging AI to consistently draw references


Retrieval Augmented Generation​

To provide more accurate responses, searching for information from outside the AI model (e.g., relevant websites, databases) and generating responses based on that


Data Formats​

JSON​

Stands for JavaScript Object Notation, a lightweight data format for storing and exchanging data


XML​

Stands for eXtensible Markup Language, used mainly to express hierarchical structure of data


CSV​

Stands for Comma-Separated Values, a format where data is stored separated by commas


Image Generation AI​

DALL-E​

An image generative AI developed by OpenAI. It generates images based on descriptions, connected with ChatGPT


Stable Diffusion​

An image generative AI developed by Stability AI Laboratory. Capable of generating high-quality images in various styles


Midjourney​

An image generative AI developed by the Midjourney Laboratory. Specialized in creating artistic images


Prompt Functions​

Implementing functions that perform specific tasks within prompts to maximize reusability

Want to learn more?

Join CodeFriends Plus membership or enroll in a course to start your journey.