• anothermember@feddit.uk
    link
    fedilink
    English
    arrow-up
    0
    arrow-down
    1
    ·
    4 days ago

    Something is wrong here, LLMs won’t spit out the same word-for-word response for the same prompt that’s not how they work.

    • ViatorOmnium@piefed.social
      link
      fedilink
      English
      arrow-up
      1
      ·
      4 days ago

      It’s probably not the same word for word, but with very similar structures. And LLMs tend to structure the text in very similar ways that don’t feel quite right.

    • hperrin@lemmy.ca
      link
      fedilink
      English
      arrow-up
      1
      ·
      4 days ago

      You are right, but you are also wrong. If they’re given the same seed, they certainly will. They are 100% deterministic. But in reality, the seed is randomly generated, so yeah, it won’t be exactly the same every time.

      • 𝙈𝙞𝙖@quokk.au
        link
        fedilink
        English
        arrow-up
        0
        ·
        4 days ago

        Even if they had the same seed, they would all need the exact same prompt. The chances of multiple people all independently coming up with the exact same prompt is highly unlikely.

        • hperrin@lemmy.ca
          link
          fedilink
          English
          arrow-up
          1
          ·
          4 days ago

          Please read the last sentence of my comment. I am not saying that the interpretation is wrong, I’m saying the statement that that’s not how LLMs work is wrong. That is how LLMs work. They are deterministic. The only reason they don’t do that in practice is because we purposefully seed them with random data to make them not do that.

        • hperrin@lemmy.ca
          link
          fedilink
          English
          arrow-up
          0
          ·
          4 days ago

          LLMs won’t spit out the same word-for-word response for the same prompt

          You can give an LLM the same seed and it will spit out the same word-for-word response. That’s how they work. It’s just a bunch of math.

          • anothermember@feddit.uk
            link
            fedilink
            English
            arrow-up
            0
            ·
            4 days ago

            You can give an LLM the same seed and it will spit out the same word-for-word response. That’s how they work. It’s just a bunch of math.

            You’re assuming that because I missed out that detail I must be ignorant of it, that’s not very charitable, I could well have been ignorant of it but you could have made your otherwise useful clarification without telling me I was wrong.

            • hperrin@lemmy.ca
              link
              fedilink
              English
              arrow-up
              1
              ·
              edit-2
              4 days ago

              You said “that’s not how they work”. But that is how they work. Same prompt = same output. Throw some random data in there to jumble things around and you get a little variance. That’s the seed, and we only need to do that because LLMs are inherently deterministic.

              Same reason Minecraft has a random seed for world generation, and block cipher algorithms use an initialization vector and/or feedback loop. We don’t want the same thing every time.

              I did say that you’re right, because the tooling we use around the LLM itself does exactly what you’re talking about. So, in practice, you’re right.