4 ms·

There will be non-trivial amounts of training data that contains sentence like "How do you spell 'apple'? a - p - p - l - e". The hyphenated or spaced out versi

There will be non-trivial amounts of training data that contains sentence like "How do you spell 'apple'? a - p - p - l - e". The hyphenated or spaced out versions are tokenized more like individual letters. So the model can learn the mapping. It's not perfect, though. There are funny failure modes where you can ask the model things like "does the word foobar have a 'b' in it" and it will say "no". (It doesn't fail in the 'foobar' case, but it does fail for some of these cases -- this was just an illustration).

hn_throwaway_99 4y ago
Thanks, yeah, this is really interesting to me. Based on your comment I tried the following test (this is the free version of ChatGPT): Me: I would like to play a would game with you. I'm going to tell you a made-up word, and then ask you a question about it. Ready? ChatGPT: Sure, I'm ready to play! Let's get started. Me: Does the made-up word plukatoon have the letter e in it? ChatGPT: No, the made-up word "plukatoon" does not have the letter "e" in it. -- Note, at this point I was really surprised it got it right. I chose "e" thinking that it would default to yes given the popularity of e. Me: Correct! Does the made-up work ziddlebum have the letter e in it? ChatGPT: No, the made-up word "ziddlebum" also does not have the letter "e" in it. -- OK, proof that it's "not perfect" as you point out Me: Does the made-up word folladel have the letter o in it? ChatGPT: Yes, the made-up word "folladel" does have the letter "o" in it. -- Edit, and finally: Me: How do you spell the made up word "platoggle"? ChatGPT: The made-up word "platoggle" is spelled as follows: P-L-A-T-O-G-G-L-E. OK, so that last one TOTALLY blew my mind. I'm guessing it's never seen the word "platoggle" before, so even if it learned previous spelling associations with other tokens, it couldn't possibly understand how to spell "platoggle" if it's treating it just as a single, never-before-seen, opaque token, could it? The more I play with this stuff the more existential crises I have.
- pimanrules 4y ago
  >it couldn't possibly understand how to spell "platoggle" if it's treating it just as a single, never-before-seen, opaque token That's not how the tokenizer works. A novel word like "platoggle" is decomposed into three separate tokens, "pl", "at", and "oggle". You can see for yourself how prompts are tokenized: https://platform.openai.com/tokenizer https://platform.openai.com/tokenizer
  - hn_throwaway_99 4y ago
    Ahh, thank you very much, definitely was missing that piece!
  - kgc 4y ago
    Why don’t they also have single letters as tokens?
    - ben_w 4y ago
      My wild guess is that if it could get things done by tokenising like that all the time, they wouldn't need to also have word-like tokens. If that is a inference time performance or training time performance or a model size issue or just total nonsense, I wouldn't know.
    - potatoinexhaust 4y ago
      They do, e.g. "gvqbkpwz" is tokenised into individual characters. Actually it was a bit tricky to construct that, since I needed to find letter combinations that are very low probability in tokeniser's training text (e.g. "gv"). So notice it doesn't contain any vowels, since almost all consonant-vowel pairs are sufficiently frequent in the training text as to be tokenised at least as a pair. E.g. "guq" is tokenised as "gu" + "q", since "gu" is common enough. (Compare "gun" which is just tokenised as a single token "gun", as it's common enough in the training set as a word on its own, so it doesn't need to tokenise it as "gu"+"n".) The only exceptions I found with consonant-vowel pairs being tokenised as pairs were ones like "qe", tokenised as "q" + "e". Or "qo" as "q"+"o". Which I guess makes sense, given these will be low-frequency pairings in the training text -- compare "qu" just tokenised as "qu". (Though I didn't test all consonant-vowel pairs, so there may be more).
- jaggederest 4y ago
  GPT4: Spell the longest state name backwards, omitting one t of your choice > The longest state name is "Massachusetts." Omitting one "t" and spelling it backwards, we get "sttesuhcassaM." I wouldn't have it quit it's day job yet. It's not great at random substring manipulation, but spelling words is so common it almost certainly has internalized some structure akin to a programmatic representation of decomposing a string into characters.
- avereveard 4y ago
  Search platoggle on Google, you may want harder made up words before going into existentialism, gpt tokenizer doesn't really work on words, and pla toggle switch are not so uncommon retail items
- alfor 4y ago
  It is intelligence, not unlike us. The biggest difference I see is that it doesn’t have consciousness: the capacity to detach it’s simulation of the world of the present. When we are conscious we jump in a simulation in our head we imagine a future where we do one thing and then we get one kind of result. If we don’t like the results we have in our simulation we jump to an alternative and so on until we find a path that we like. We can attach ours simulation to past memories and again simulate alternative, to try to integrate what happened to us. GPT doesn’t seem to have that yet, it’s always in the present moment, like a dog, no worry for the future.
gl-prod 4y ago
Interestingly, sometimes it fails like this: "Yes, the second letter of foobar is b. The word foobar is spelled as f-o-o-b-a-r. Do you know what foobar means?"

hn_throwaway_99 4y ago

Thanks, yeah, this is really interesting to me. Based on your comment I tried the following test (this is the free version of ChatGPT): Me: I would like to play a would game with you. I'm going to tell you a made-up word, and then ask you a question about it. Ready? ChatGPT: Sure, I'm ready to play! Let's get started. Me: Does the made-up word plukatoon have the letter e in it? ChatGPT: No, the made-up word "plukatoon" does not have the letter "e" in it. -- Note, at this point I was really surprised it got it right. I chose "e" thinking that it would default to yes given the popularity of e. Me: Correct! Does the made-up work ziddlebum have the letter e in it? ChatGPT: No, the made-up word "ziddlebum" also does not have the letter "e" in it. -- OK, proof that it's "not perfect" as you point out Me: Does the made-up word folladel have the letter o in it? ChatGPT: Yes, the made-up word "folladel" does have the letter "o" in it. -- Edit, and finally: Me: How do you spell the made up word "platoggle"? ChatGPT: The made-up word "platoggle" is spelled as follows: P-L-A-T-O-G-G-L-E. OK, so that last one TOTALLY blew my mind. I'm guessing it's never seen the word "platoggle" before, so even if it learned previous spelling associations with other tokens, it couldn't possibly understand how to spell "platoggle" if it's treating it just as a single, never-before-seen, opaque token, could it? The more I play with this stuff the more existential crises I have.

pimanrules 4y ago
>it couldn't possibly understand how to spell "platoggle" if it's treating it just as a single, never-before-seen, opaque token That's not how the tokenizer works. A novel word like "platoggle" is decomposed into three separate tokens, "pl", "at", and "oggle". You can see for yourself how prompts are tokenized: https://platform.openai.com/tokenizer https://platform.openai.com/tokenizer
- hn_throwaway_99 4y ago
  Ahh, thank you very much, definitely was missing that piece!
- kgc 4y ago
  Why don’t they also have single letters as tokens?
  - ben_w 4y ago
    My wild guess is that if it could get things done by tokenising like that all the time, they wouldn't need to also have word-like tokens. If that is a inference time performance or training time performance or a model size issue or just total nonsense, I wouldn't know.
  - potatoinexhaust 4y ago
    They do, e.g. "gvqbkpwz" is tokenised into individual characters. Actually it was a bit tricky to construct that, since I needed to find letter combinations that are very low probability in tokeniser's training text (e.g. "gv"). So notice it doesn't contain any vowels, since almost all consonant-vowel pairs are sufficiently frequent in the training text as to be tokenised at least as a pair. E.g. "guq" is tokenised as "gu" + "q", since "gu" is common enough. (Compare "gun" which is just tokenised as a single token "gun", as it's common enough in the training set as a word on its own, so it doesn't need to tokenise it as "gu"+"n".) The only exceptions I found with consonant-vowel pairs being tokenised as pairs were ones like "qe", tokenised as "q" + "e". Or "qo" as "q"+"o". Which I guess makes sense, given these will be low-frequency pairings in the training text -- compare "qu" just tokenised as "qu". (Though I didn't test all consonant-vowel pairs, so there may be more).
jaggederest 4y ago
GPT4: Spell the longest state name backwards, omitting one t of your choice > The longest state name is "Massachusetts." Omitting one "t" and spelling it backwards, we get "sttesuhcassaM." I wouldn't have it quit it's day job yet. It's not great at random substring manipulation, but spelling words is so common it almost certainly has internalized some structure akin to a programmatic representation of decomposing a string into characters.
avereveard 4y ago
Search platoggle on Google, you may want harder made up words before going into existentialism, gpt tokenizer doesn't really work on words, and pla toggle switch are not so uncommon retail items
alfor 4y ago
It is intelligence, not unlike us. The biggest difference I see is that it doesn’t have consciousness: the capacity to detach it’s simulation of the world of the present. When we are conscious we jump in a simulation in our head we imagine a future where we do one thing and then we get one kind of result. If we don’t like the results we have in our simulation we jump to an alternative and so on until we find a path that we like. We can attach ours simulation to past memories and again simulate alternative, to try to integrate what happened to us. GPT doesn’t seem to have that yet, it’s always in the present moment, like a dog, no worry for the future.

gl-prod 4y ago

Interestingly, sometimes it fails like this: "Yes, the second letter of foobar is b. The word foobar is spelled as f-o-o-b-a-r. Do you know what foobar means?"