Learn PromptingRepetition
Repeating the same word within a prompt, or similar phrases can cause the model to emphasize that word in the generated image. For example, @Phillip Isola generated these waterfalls with DALLE:
A beautiful painting of a mountain next to a waterfall..
A very very very very very very very very very very very very very very very very very very very very very very beautiful painting of a mountain next to a waterfall.
The emphasis on the word very seems to improve generation quality! Repetition can
also be used to emphasize subject terms. For example, if you want to generate an image
of a planet with aliens, using the prompt A planet with aliens aliens aliens aliens aliens aliens aliens aliens aliens aliens aliens aliens
will make it more likely that aliens are in the resultant image. The following images are made with Stable Diffusion.
A planet with aliens
A planet with aliens aliens aliens aliens aliens aliens aliens aliens aliens aliens aliens aliens
Notes
This method is not perfect, and using weights (next article) is often a better option.
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.
Footnotes
-
Oppenlaender, J. (2022). A Taxonomy of Prompt Modifiers for Text-To-Image Generation. ↩