Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Teaching GPT-3 to reverse words

twitter.com

Post reply on HN
Print view View original

21–30 of 81 posts

Re: Teaching GPT-3 to reverse words

#21

Post by visarga »

oneepic wrote: post #17
jameshart wrote: post #15

Oh, I’m so looking forward to my next coding interview. “Okay, could you show me on the whiteboard how you might go about writing a program that can reverse a string?” “Great, so I’m going to start by initializing a simple transformer-based neural network with 175 billion parameters and 96 attention layers, and I’m going to train it on a corpus of 45 terabytes of data tokenized into about 500 billion tokens…”

"Cool, so what do you think would be the time complexity of that? Do you think we can maybe do better than that?"

Not if you also want a short poem where each word starts with a letter from the original word, and then a short literary commentary on it.

Re: Teaching GPT-3 to reverse words

#22

Post by visarga »

mberning wrote: post #16
swid wrote: post #12

Earlier quoted context omitted.

I meant it mostly as a joke, but there is a certain amount of irony to it. This goes way beyond prompt engineering - he wrote an algorithm to run on GPT in a way you would not expect a non-programmer to write. I think the idea is cool and the process to write it was revealing.

Right. What non-programmer is going to think to turn a word into character list with positional metadata sprinkled in.

I used a similar technique for a completely unrelated task. My "original" idea.

Re: Teaching GPT-3 to reverse words

#23

Post by a65cec93b »

> GPT-3 correctly reverses long words! But to get there, we had to teach GPT-3 the algorithm to use to get around its limitations.

Has GPT-3 really been "taught" anything here? If you don't provide an explicit example as the context of your input, GPT-3 does not retain the ability to reverse words.

Re: Teaching GPT-3 to reverse words

#24

Post by jameshart »

oneepic wrote: post #17
jameshart wrote: post #15

Oh, I’m so looking forward to my next coding interview. “Okay, could you show me on the whiteboard how you might go about writing a program that can reverse a string?” “Great, so I’m going to start by initializing a simple transformer-based neural network with 175 billion parameters and 96 attention layers, and I’m going to train it on a corpus of 45 terabytes of data tokenized into about 500 billion tokens…”

"Cool, so what do you think would be the time complexity of that? Do you think we can maybe do better than that?"

Actually it turns out it's O(n). Which goes to show that constant factors can be more important than you think when looking at raw time complexity big-O.

Re: Teaching GPT-3 to reverse words

#25

Post by f38zf5vdt »

a65cec93b wrote: post #23

> GPT-3 correctly reverses long words! But to get there, we had to teach GPT-3 the algorithm to use to get around its limitations. Has GPT-3 really been "taught" anything here? If you don't provide an explicit example as the context of your input, GPT-3 does not retain the ability to reverse words.

No, it isn't taught anything. GPT3 text generation is effectively a really fancy autocompletion algorithm based on the n-many previous tokens in a rolling window. You can only "teach" GPT3 something within that window, and it doesn't "learn" there, it just tries its best to generate content based on what is stored in its massive n-dimension table of graph edges for tokens.

That is also why it has such a strong propensity to lose the plot once you are outside of that window size and it's generating new content based on self-generated content.

Re: Teaching GPT-3 to reverse words

#26

Post by gwern »

6gvONxR4sf7o wrote: post #11
mrfusion wrote: post #2

> Tokens are chunks of characters. For example, the word “alphabet” gets broken up into the tokens “alph" and "abet”. I didn’t know that. Seems like it would confuse it during training. Anyone able to explain?

The alternatives are learning at the character level (way more complex, and scales badly in memory/compute), or learning at the whole word level (needs absurdly massive dictionary of words, and still can’t handle really rare/novel words). Breaking things into a set of subwords that allows you to encode any string solves lots of problems and is the relatively standard way to do things these days.

> The alternatives are learning at the character level (way more complex

No, BPEs are more complex: you have a whole additional layer of preprocessing, with all sorts of strange and counterintuitive downstream effects and brand new ways to screw up (fun quiz question: everyone knows that BPEs use '' tokens to denote document breaks; what does the string '' encode to?). BPEs are reliably one of the ways that OA API users screw up, especially when trying to work with longer completions or context windows.

But a character is a character.

> and scales badly in memory/compute)

Actually very competitive: https://arxiv.org/abs/2105.13626#google (Especially if you account for all the time and effort and subtle bugs caused by BPEs.)

Re: Teaching GPT-3 to reverse words

#27

Post by SemanticStrengh »

mrfusion wrote: post #2

> Tokens are chunks of characters. For example, the word “alphabet” gets broken up into the tokens “alph" and "abet”. I didn’t know that. Seems like it would confuse it during training. Anyone able to explain?

This is masked token learning, which is used e.g by BERT. This is obscolete and alternatives such as XLNET are much superior but there is too much inertia in the industry and newer large models are still built with the same lossy encoding..

Re: Teaching GPT-3 to reverse words

#28

Post by SemanticStrengh »

aeternum wrote: post #5
mrfusion wrote: post #2

> Tokens are chunks of characters. For example, the word “alphabet” gets broken up into the tokens “alph" and "abet”. I didn’t know that. Seems like it would confuse it during training. Anyone able to explain?

Humans also think about words in terms of subcomponents, languages make heavy use of prefixes and suffixes for example.

This is not the same.. The masks are randomized and lossy. Although yes there is potential for a transformer specially trained to segment prefixes/affixes/suffixes, it might augment some of its encoding abilities, see e.g spanbert for a related example of opportunity.

Re: Teaching GPT-3 to reverse words

#29

Post by yunyu »

f38zf5vdt wrote: post #25
a65cec93b wrote: post #23

> GPT-3 correctly reverses long words! But to get there, we had to teach GPT-3 the algorithm to use to get around its limitations. Has GPT-3 really been "taught" anything here? If you don't provide an explicit example as the context of your input, GPT-3 does not retain the ability to reverse words.

No, it isn't taught anything. GPT3 text generation is effectively a really fancy autocompletion algorithm based on the n-many previous tokens in a rolling window. You can only "teach" GPT3 something within that window, and it doesn't "learn" there, it just tries its best to generate content based on what is stored in its massive n-dimension table of graph edges for tokens. That is also why it has such a strong propen…

You can update the "graph edges" with content longer than the window by fine tuning: https://beta.openai.com/docs/guides/fine-tuning
Post reply on HN