Learn PromptingWeighted Terms
- Weighting Terms allows you explicitly adjust the emphasis on a given feature of your image, in either direction.
What are Weighted Terms?
Some models (Stable Diffusion, Midjourney, etc.) allow you to assign weights to certain terms in a prompt. Such weighted terms can be used to emphasize certain words or phrases in the generated image. It can also be used to de-emphasize certain words or phrases in the generated image. Let's consider a simple example:
An Example of Weighted Terms in Image Prompts
Here are a few mountains generated by Stable Diffusion, with the prompt mountain.
However, if we want mountains without trees, we can use the prompt mountain | tree:-10. Since we weighted trees very negatively, they do not appear in the generated image.
Weighted terms can be combined into more complicated prompts, like
A planet in space:10 | bursting with color red, blue, and purple:4 | aliens:-10 | 4K, high quality
Conclusion
Weighted terms allow you to explicitly tell the model which aspects of the output image or more or less important, giving you even more control in specifying your image prompts.
FAQ
How can weighted terms help my image prompts?
Weighted terms are a capability of certain models that allow you to explicitly define the weights of certain words or phrases in an input. In this way, you can emphasize or de-emphasize the things you'd like portrayed in the AI-generated image.
Notes
Read more about weighting in some of the resources at the end of this chapter.
Sander Schulhoff
Sander Schulhoff is the CEO of HackAPrompt and Learn Prompting. He created the first Prompt Engineering guide on the internet, two months before ChatGPT was released, which has taught 3 million people how to prompt ChatGPT. He also partnered with OpenAI to run the first AI Red Teaming competition, HackAPrompt, which was 2x larger than the White House's subsequent AI Red Teaming competition. Today, HackAPrompt partners with the Frontier AI labs to produce research that makes their models more secure. Sander's background is in Natural Language Processing and deep reinforcement learning. He recently led the team behind The Prompt Report, the most comprehensive study of prompt engineering ever done. This 76-page survey, co-authored with OpenAI, Microsoft, Google, Princeton, Stanford, and other leading institutions, analyzed 1,500+ academic papers and covered 200+ prompting techniques.