5 ms·
They memorize the answers not the process to arrive at answers
by fsmv 1y ago
They memorize the answers not the process to arrive at answers
- IshKebab 1y agoThis has been disproven so many times... They clearly do both. You can trivially prove this yourself.
- 0xWTF 1y ago> You can trivially prove this yourself. Given the long list of dead philosophers of mind, if you have a trivial proof, would you mind providing a link?
- pdabbadabba 1y agoIt’s really easy: go to Claude and ask it a novel question. It will generally reason its way to a perfectly good answer even if there is no direct example of it in the training data.
- MichaelZuo 1y agoHow do you know it’s a novel question?
- IshKebab 1y agoIt's not exactly difficult to come up with a question that's so unusual the chance of it being in the training set is effectively zero.
- troupo 1y agoAnd as any programmer will tell you: they immediately devolve into "hallucinating" answers, not trying to actually reason about the world. Because that's what they do: they create statistically plausible answers even if those answers are complete nonsense.
- MichaelZuo 1y agoCan you provide some examples of these genuinely unique questions?
- pdabbadabba 1y agoI'm not sure what you mean by "genuinely." But in the coding context LLMs answer novel questions all the time. My codebase uses components and follows patterns that an LLM will have seen before, but the actual codebase is unique. Yet, the LLM can provide detailed explanations about how it works, what bugs or vulnerabilities it might have, modify it, or add features to it.
- MichaelZuo 1y agoIt must not have existed prior in any text database whatsoever.
- pdabbadabba 1y agoIt certainly wasn't. The codebase is thousands of lines of bespoke code that I just wrote.
- drw85 1y agoWhich pretty much every line in it was written similarly somewhere else before, including an explanation and is somehow included in the massive data set it was trained on. So far i have asked the AI some novel questions and it came up with novel answers full of hallucinated nonsense, since it copied some similarly named setting or library function and replaced a part of it's name with something i was looking for.
- pdabbadabba 1y agoAnd this training data somehow includes an explanation of how these individual lines (with variable names unique to my application) work together in my unique combination to produce a very specific result? I don't buy it. And... > pretty much Is it "pretty much" or "all"? The claim that the LLM simply has simply memorized all of its responses seems to require "all."
- hackinthebochs 1y agoYou have probably seen examples of LLMs doing the "mirror test", i.e. identifying themselves in screenshots and referring to the screenshot from the first person. That is a genuinely novel question as an "LLM mirror test" wasn't a concept that existed before about a year ago.
- MichaelZuo 1y agoElephant mirror tests existed, so it doesn’t seem all that novel when the word “elephant” could just be substituted for the word “LLM”?
- hackinthebochs 1y agoThe question isn't about universal novelty, but whether the prompt/context is novel enough such that the LLM answering competently demonstrates understanding. The claim of parroting is that the dataset contains a near exact duplicate of any prompt and so the LLM demonstrating what appears to be competence is really just memorization. But if an LLM can generalize from an elephant mirror test to an LLM mirror test in an entirely new context (showing pictures and being asked to describe it), that demonstrates sufficient generalization to "understand" the concept of a mirror test.
- MichaelZuo 1y agoHow do you know it’s the one generalizing? Likely there has been at least one text that already does that for say dolphin mirror tests or chimpanzee mirror teats.
- keerthiko 1y agoWhen LLM's come up with answers to questions that aren't directly exampled in the training data, that's not proof at all that it reasoned its way there — it can very much still be pattern matching without insight from the actual code execution of the answer generation. If we were taking a walk and you asked me for an explanation for a mathematical concept I have not actually studied, I am fully capable of hazarding a casual guess based on the other topics I have studied within seconds. This is the default approach of an LLM, except with much greater breadth and recall of studied topics than I, as a human, have. This would be very different than if we sat down at a library and I applied the various concepts and theorems I already knew to make inferences, built upon them, and then derived an understanding based on reasoning of the steps I took (often after backtracking from several reasoning dead ends) before providing the explanation. If you ask an LLM to explain their reasoning, it's unclear whether it just guessed the explanation and reasoning too, or if that was actually the set of steps it took to get to the first answer they gave you. This is why LLMs are able to correct themselves after claiming strawberry has 2 rs, but when providing (guessing again) their explanations they make more "relevant" guesses.
- pdabbadabba 1y agoI'm not sure what "just guessed" means here. My experience with LLMs is that their "guesses" are far more reliable than a human's casual guess. And, as you say, they can provide cogent "explanations" of their "reasoning." Again, you say they might be "just guessing" at the explanation, what does that really mean if the explanation is cogent and seems to provide at least a plausible explanation for the behavior? (By the way, I'm sure you know that plenty of people think that human explanations for their behavior are also mere narrative reconstructions.) I don't have a strong view about whether LLMS are really reasoning -- whatever that might mean. But the point I was responding to is that LLMS have simply memorized all the answers. That is clearly not true under any normal meanings of those words.
- IshKebab 1y agoLLMs clearly don't reason in the same way that humans or SMT solvers do. That doesn't mean they aren't reasoning.
- IshKebab 1y agoJust go and ask ChatGPT or Claude something that can't possibly be in its training set. Make something up. If it is only memorising answers then it will be impossible for it to get the correct result. A simple nonsense programming task would suffice. For example "write a Python function to erase every character from a string unless either of its adjacent characters are also adjacent to it in the alphabet. The string only contains lowercase a-z" That task isn't anywhere in its training set so they can't memorise the answer. But I bet ChatGPT and Claude can still do it. Honestly this is sooooo obvious to anyone that has used these tools, it's really insane that people are still parroting (heh) the "it just memorises" line.
- troupo 1y agoPeople who say that LLMs memorize stuff are just as clueless who assume that there's any reasoning happening. They generate statistically plausible answers (to simplify the answer) based on the training set and weights they have.
- Tijdreiziger 1y agoWhat if that’s all we’re doing, though?
- troupo 1y agoMost of us definitely do :) Or we do it most of the time :)
- imiric 1y agoLLMs don't "memorize" concepts like humans do. They generate output based on token patterns in their training data. So instead of having to be trained on every possible problem, they can still generate output that solves it by referencing the most probable combination of tokens for the specified input tokens. To humans this seems like they're truly solving novel problems, but it's merely a trick of statistics. These tools can reference and generate patterns that no human ever could. This is what makes them useful and powerful, but I would argue not intelligent.
- EternalFury 1y agoThey learn the value of specific actions in specific contexts based on the rewards they received during their play time. Specific actions and specific contexts are not transferable for various reasons. John quoted that varying frame rates and variable latency between action and effect really confuse the models.