What can LLMs never do? • Hacker News Board

Re: What can LLMs never do?

#141

Post by layer8 » Sat, Apr 27, 2024, 10:11 PM UTC

dragonwriter wrote: post #120

cs702 wrote: post #14

Fantastic essay. Highly recommended! I agree with all key points: * There are problems that are easy for human beings but hard for current LLMs (and maybe impossible for them; no one knows). Examples include playing Wordle and predicting cellular automata (including Turing-complete ones like Rule 110). We don't fully understand why current LLMs are bad at these tasks. * Providing an LLM with examples and step-by-step…

> There are problems that are easy for human beings but hard for current LLMs (and maybe impossible for them; no one knows). Examples include playing Wordle and predicting cellular automata (including Turing-complete ones like Rule 110). We don’t fully understand why current LLMs are bad at these tasks. I thought we did know for things like playing Wordle, that its because they deal with words as sequence of tokens t…

Regarding Wordle, it should be straightforward to make a token-based version of it, and I would assume that that has been tried. It seems the obvious thing to do when one is interested in the reasoning abilities necessary for Wordle.

Re: What can LLMs never do?

#142

Post by bboygravity » Sat, Apr 27, 2024, 10:14 PM UTC

LLMs can never experiment with/in the real world to find answers to open questions.

That's the summary of "The book of Why" in one sentence as I understand it.

Re: What can LLMs never do?

#143

Post by layer8 » Sat, Apr 27, 2024, 10:15 PM UTC

j16sdiz wrote: post #125

_wire_ wrote: post #119

Earlier quoted context omitted.

>We don't fully understand why current LLMs are bad at these tasks. In complete seriousness, can anyone can explain why LLMs are good at some tasks?

> why LLMs are good at some tasks? Like how we explain human doing tasks -- they are evolved to do that. I believe this is a non-answer, but if we are satisfied with that non answer for human, why not LLMs?

I would argue that we are not satisfied with that answer for humans either.

Re: What can LLMs never do?

#144

Post by jiggawatts » Sat, Apr 27, 2024, 10:17 PM UTC

bxparks wrote: post #47

Xenoamorphous wrote: post #21

Earlier quoted context omitted.

“Ten” is a word, “10” are digits. I’m not a native English speaker, how would you write it? FWIW the LLMs get it right many times, but fail other times.

I couldn't understand the original wording either, but after reading one of the sibling comments that explains it, it suddenly made sense. I think you left out a few words that most English writers would include. So instead of: > "ensure that numbers from one to ten as written as words and numbers greater ten as digits in the given text", something like the following might be better for most people: > "ensure that th…

Hilariously, you can ask GPT 4 to explain the “why” of arbitrary grammar fixes.

Re: What can LLMs never do?

#145

Post by richardw » Sat, Apr 27, 2024, 10:19 PM UTC

_wire_ wrote: post #119

cs702 wrote: post #14

Fantastic essay. Highly recommended! I agree with all key points: * There are problems that are easy for human beings but hard for current LLMs (and maybe impossible for them; no one knows). Examples include playing Wordle and predicting cellular automata (including Turing-complete ones like Rule 110). We don't fully understand why current LLMs are bad at these tasks. * Providing an LLM with examples and step-by-step…

>We don't fully understand why current LLMs are bad at these tasks. In complete seriousness, can anyone can explain why LLMs are good at some tasks?

LLM’s are a compressed and lossy form of our combined writing output, which it turns out is similarly structured enough to make new combinations of text seem reasonable, even enough to display simple reasoning. I find it useful to think “what can I expect from speaking with the dataset of combined writing of people”, rather than treating a basic LLM as a mind.

That doesn’t mean we won’t end up approximating one eventually, but it’s going to take a lot of real human thinking first. For example, ChatGPT writes code to solve some questions rather than reasoning about it from text. The LLM is not doing the heavy lifting in that case.

Give it (some) 3D questions or anything where there isn’t massive textual datasets and you often need to break out to specialised code.

Another thought I find useful is that it considers its job done when it’s produced enough reasonable tokens, not when it’s actually solved a problem. You and I would continue to ponder the edge cases. It’s just happy if there are 1000 tokens that look approximately like its dataset. Agents make that a bit smarter but they’re still limited by the goal of being happy when each has produced the required token quota, missing eg implications that we’d see instantly. Obviously we’re smart enough to keep filling those gaps.

Re: What can LLMs never do?

#146

Post by layer8 » Sat, Apr 27, 2024, 10:20 PM UTC

JeremyHerrman wrote: post #122

Why can't we resolve this with synthetic data? Just take the original statements and ask another LLM to flip around the relation, then train on both. I tried this with GPT-4 and it seems to understand the task [0]: Me: For each of the following relational statements, flip the relation around and generate a new statement or question answer pair. Input: Bluey's father is Bandit Output: Bandit's daughter is Bluey Input:…

The difficulty would likely be that the “X is Y” relations don’t just occur in that simple, separated-out form, but in a whole spectrum of more complicated contexts, and the latter probably dominate. You would have to identify those and synthesize a text corpus that contains the reverse notions across a similar spectrum.

Re: What can LLMs never do?

#147

Post by dragonwriter » Sat, Apr 27, 2024, 10:21 PM UTC

dcchambers wrote: post #7

Guarantee an output will be consistent every time.

> Guarantee an output will be consistent every time.

If you mean “consistent with a prior run with the same input”, LLMs can absolutely do that, though for most purposes settings are chosen deliberately which do not.

If you mean “consistent with some external structural specification”, LLMs can do that, too, e.g., via grammar specifications.

Re: What can LLMs never do?

#148

Post by lukeschlather » Sat, Apr 27, 2024, 10:23 PM UTC

logtempo wrote: post #30

vsnf wrote: post #20

Earlier quoted context omitted.

> ensure that numbers from one to ten as written as words and numbers greater ten as digits in the given text I can’t fault llms for not knowing what to do here because I, a human, have no idea what on earth this means.

I think he mean that numbers less or equal than ten are written as words, and others are written as numbers. Given the many reaponses, it would be fun to aee if llm beat humans on understanding the sentence ahah

to me the main problem is that it should read "numbers greater than ten." I asked Gemini to rephrase it and Gemini produced correct English with the intended meaning:

> Change all numbers between one and ten to words, and write numbers eleven and above as digits in the text.

It even used eleven rather than ten which sounds like counting.

Re: What can LLMs never do?

#149

Post by danenania » Sat, Apr 27, 2024, 10:35 PM UTC

SomeCallMeTim wrote: post #136

_wire_ wrote: post #119

Earlier quoted context omitted.

>We don't fully understand why current LLMs are bad at these tasks. In complete seriousness, can anyone can explain why LLMs are good at some tasks?

LLMs are good at tasks that don't require actual understanding of the topic. They can come up with excellent (or excellent-looking-but-wrong) answers to any question that their training corpus covers. In a gross oversimplification, the "reasoning" they do is really just parroting a weighted average (with randomness injected) of the matching training data. What they're doing doesn't really match any definition of "und…

> the "reasoning" they do is really just parroting a weighted average (with randomness injected) of the matching training data

Perhaps our brains are doing exactly the same, just with more sophistication?

Re: What can LLMs never do?

#150

Post by eqmvii » Sat, Apr 27, 2024, 10:42 PM UTC

I just asked MetaAI to help me with Wordle and it understood and gave me a sane answer, so…

Edit: ah, I spoke too soon. My first question was too "easy" but I asked a few more, and sure enough... it can understand what I'm asking and it can write an answer that's well formed, but it's fundamentally not understanding the rules of the game or giving me valid guesses. Cute!