English
Share
Sign In
Introducing Prompts: Leveraging what language models do well
The work of improving prompts is a field that is developing in various research fields to maximize the efficiency of language models (LM). This field plays an important role in unraveling the complexity of large-scale language models (LLM) and revealing their capabilities and limitations. To be more honest, you can think of it as a technique used to maximize the efficiency of limited computing resources and limited language models, and a way to draw out the potential of existing models.
Researchers and developers are increasingly using prompting to improve LLM performance in a variety of tasks, from question answering to arithmetic reasoning. They are designing sophisticated and efficient prompting techniques to make the interface between LLM and various digital tools more robust.
This guide is designed to introduce you to the world of prompts and provide a basic understanding of how to create prompts to effectively interact and guide your LLM. Understanding prompts will help you understand not only the 'how' of these interactions, but also the 'why' of them.
OpenAI Playground is a space where anyone can experiment with various language models and test products provided by OpenAI. If you are using chatGPT, you can log in with your original ID. If you do not have an ID, we recommend creating one. It is free to use.
As mentioned earlier, all examples are conducted on chatGPT's free GPT-3.5 model. For those using GPT-4, please note that many parts have been resolved through performance, so the effect of prompt engineering is minimal.
ⓒ 2023. Haebom, all rights reserved.
It may be used for commercial purposes with permission from the copyright holder, provided the source is cited.