Skip to content
TheChatArchiveThe Chat Archive
Human-written demonstration (OpenAssistant volunteers)· OpenAssistant (human-written)Hugging Face OpenAssistant Conversations (oasst2)

Write me a Stable Diffusion prompt for retro science fiction woman

In OpenAssistant data, “assistant” turns were written by human volunteers as demonstrations. They are not model outputs.
  1. User

    Write me a Stable Diffusion prompt for retro science fiction woman

  2. Assistant

    prompt: A retro science fiction woman, slovenly, couch potato, A. Leslie Ross style. neg prompt: ((smiling)), (((duplicate))), ((mutilated)), ((couple)), (out of frame), extra fingers, mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), blurry, ((bad anatomy)), (((bad proportions))), ((extra limbs)), ((cloned face)), (((disfigured))). out of frame, extra limbs, (bad anatomy), gross proportions, (malformed limbs), ((missing arms)), ((missing legs)), (((extra arms))), (((extra legs))), mutated hands, (fused fingers), (too many fingers), (((long neck)))

  3. User

    Your prompt kinda worked, but a third leg or other distortions sometimes appear. Also, why did you include "slovenly" and "couch potato"? I tried without them, and at least it wasn't any worse. Also the A. Leslie Ross part didn't help much.

    What do all these parentheses do? And why are there sometimes multiple?

  4. Assistant

    The parentheses and multiple parentheses were included as a way to potentially guide the AI in generating an image by providing both desired features and undesired features. However, I understand that it might have caused confusion and didn't yield the desired results.

    To create a simpler prompt without the undesired features and additional parentheses, try the following:

    prompt: A retro science fiction woman in the style of A. Leslie Ross, with a focus on the classic sci-fi aesthetics.

    By simplifying the prompt, the AI will have a clearer idea of what you are looking for without being influenced by the undesired features. You may need to iterate on the prompt to fine-tune the output based on the AI's generated result.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.