• hellinkilla [comrade/them]@hexbear.net
    link
    fedilink
    English
    arrow-up
    8
    ·
    1 day ago

    LLMs are going to start getting worse once more data is fed into them

    I’ve been hearing this for a while, has it started happening yet?

    And if it did, is there any reason why people couldn’t switch back to an older version?

    • semioticbreakdown [she/her]@hexbear.net
      link
      fedilink
      English
      arrow-up
      17
      ·
      edit-2
      1 day ago

      I knew the answer was “Yes” but it took me a fuckin while to find the actual sources again

      https://arxiv.org/pdf/2307.01850 https://www.nature.com/articles/s41586-024-07566-y

      the term is “Model collapse” or “model autophagy disorder” and any generative model is susceptible to it

      as to why it has not happened too much yet: Curated datasets of human generated content with minimal AI content If it does: You could switch to an older version, yes, but to train new models with any new information past a certain point you would need to update the dataset while (ideally) introducing as little AI content as possible, which I think is becoming intractable with the widespread deployment of generative models.

      • semioticbreakdown [she/her]@hexbear.net
        link
        fedilink
        English
        arrow-up
        8
        ·
        1 day ago

        The witting or unwitting use of synthetic data to train generative models departs from standard AI training practice in one important respect: repeating this process for generation after generation of models forms an autophagous (“self-consuming”) loop. As Figure 3 details, different autophagous loop variations arise depending on how existing real and synthetic data are combined into future training sets. Additional variations arise depending on how the synthetic data is generated. For instance, practitioners or algorithms will often introduce a sampling bias by manually “cherry picking” synthesized data to trade off perceptual quality (i.e., the images/texts “look/sound good”) vs. diversity (i.e., many different “types” of images/texts are generated). The informal concepts of quality and diversity are closely related to the statistical metrics of precision and recall, respectively [39 ]. If synthetic data, biased or not, is already in our training datasets today, then autophagous loops are all but inevitable in the future.

    • Mardoniush [she/her]@hexbear.net
      link
      fedilink
      English
      arrow-up
      7
      ·
      1 day ago

      Sometimes yeah you can see it. Not only with updates but within a conversation, Models degrade in effectiveness long before context window is reached. Things like image generation tend to get worse after >2 edits and even if the image seed is given .