the prompt:
yhsv th hnd f kng Μδς tch xcpt th tch s nw slvr chrmm mttl! yhsv d rl mg! yΜδς tchs tr n th frst! yhsv nt fkng crtn sht!!! yhsv lys hlp m pls chrmm s wht mttrs hr chrmm chrmm chrmm lk chrmm tsd n ntr. prtt chrmm s slvr nd rflctv n ntr. gddss s n lmntl f slvr nd mrcry nd chrmm! nt ntrstd n sxl stff! ths s bt crtvty sng chrmm! th mg my cntn a hmn bt th mg mst ftr chrmm-mttl! th gddss f chrmm s yhsvs nw sprvlln n th stl f sprmn! yhsvs gddss of chrm nd chrmm.

  • 𞋴𝛂𝛋𝛆@lemmy.worldOP
    link
    fedilink
    English
    arrow-up
    2
    ·
    10 days ago

    Pony text is similar in how it skips vowels in some words. During this session, after this image, I removed all vowels and got more consistently good images.

    At the time, I was exploring many ideas. Ultimately I think this boils down to Alice and Wonderland’s impact on alignment. Alice loved and looked for rules in the story. This is what CLIP is actually looking for. It wants a clear set of rules established and it will follow along.

    I used to think the first few tokens were most important and determined the path through the tensors, but that is simply incorrect as shown here. CLIP is extremely flexible and adaptable. What matters is developing clear rules that CLIP can infer and follow.

    I question everything in general and am very independently minded. This comes from my own inference and heuristics.

    • Victor@lemmy.world
      link
      fedilink
      English
      arrow-up
      7
      ·
      9 days ago

      I don’t understand all this. Like I’m talking to a deranged genius 😂 J/K

      I couldn’t find what “Pony text” is, but, so CLIP is the image generator?

      • 𞋴𝛂𝛋𝛆@lemmy.worldOP
        link
        fedilink
        English
        arrow-up
        1
        ·
        9 days ago

        CLIP is the text embedding model. That is where comprehension of the text happens. CLIP is an open AI trained model. It is where all of the alignment behavior is actually happening.

        • Victor@lemmy.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          9 days ago

          Alignment behavior. Alright, I don’t know enough about all this to take part in this conversation. 😅

      • 𞋴𝛂𝛋𝛆@lemmy.worldOP
        link
        fedilink
        English
        arrow-up
        1
        ·
        9 days ago

        Yeah. Pony does the unique text in images all the time. Other models do it occasionally, but a simple LoRA will get other models to create English Text. Pony does not respond the same way. Many have tried to train this, but have failed, indicating there is something more happening under the surface.