this post was submitted on 18 Jun 2023
44 points (100.0% liked)
Programmer Humor
32483 readers
453 users here now
Post funny things about programming here! (Or just rant about your favourite programming language.)
Rules:
- Posts must be relevant to programming, programmers, or computer science.
- No NSFW content.
- Jokes must be in good taste. No hate speech, bigotry, etc.
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Fun fact: If you google those codes you find out that they are "real" codes, but they don't actually activate Windows. I think they are something that are used as placeholders in the upgrade from Windows 8 to 10 or something, but don't know the specifics.
ChatGPT actually can't create new "words", just regurgitate words that it's seen somewhere before!
Sure it can create new words. It can't create new tokens, would be more correct, I think. But a token is just a text fragment, and, as far as I know, they can range from being several words to being single characters.
I got it it say vilumplox. It doesn't return any Google search results.
Yep yep, statistical analysis as to the frequency of tokens in the training text.
Brand new, never-before-seen Windows keys have a frequency of zero occurrences per billion words of training data.
That isn't actually what's important. It's the frequency of the token, which could be as simple as single characters. The frequency of those is certainly not zero.
LLMs absolutely can make up new words, word combinations, or sentences.
That's not to say chatgpt can actually give you good windows keys, but it isn't a fundamental limitation of LLMs.
Okay, I'll take your word for it.
I've never ever, in many hours of playing with ChatGPT as a toy, had it make up a word. Hallucinate wildly, yes, but not stogulate a word out of nothing.
I'd love to know more, though. How does it combine new words? Do you have any examples of words ChatGPT has made up? This is fascinating to me, as it means the model is much less chained to the training data than I thought.
A lot of compound words are actually multiple tokens so there's nothing stopping the LLM from generating the tokens in a new order thereby creating a new word.