6 ms·
>For example, we loaded the entire text of The Great Gatsby into Claude-Instant (72K tokens) and modified one line to say Mr. Carraway was “a software engineer
by Imnimo 3y ago
>For example, we loaded the entire text of The Great Gatsby into Claude-Instant (72K tokens) and modified one line to say Mr. Carraway was “a software engineer that works on machine learning tooling at Anthropic.” When we asked the model to spot what was different, it responded with the correct answer in 22 seconds.
This sort of needle-in-a-haystack retrieval is definitely impressive, and it makes a lot more sense to achieve this in-context rather than trying to use a vector database if you can afford it.
I'm curious, though, whether there are diminishing returns in terms of how much analysis the model can do over those 100k tokens in a single forward pass. A human reading modified-Gatsby might eventually spot the altered line, but they'd also be able to answer questions about the overarching plot and themes of the novel, including ones that cannot be deduced from just a small number of salient snippets.
I'd be curious to see whether huge-context models are also able to do this, or if they start to have trouble when the bottleneck becomes reasoning capacity rather than input length. I feel like it's hard to predict one way or the other without trying it, just because LLMs have already demonstrated a lot of surprising powers.
- tunesmith 3y agoI've been curious about this for a while, I have a hobby use-case of wanting to input in-progress novellas and then asking it questions about plot holes, open plot threads, and if new chapter "x" presents any serious plot contradiction problems. I haven't tried exploring that with a vectordb-embeddings approach yet.
- make3 3y agoThis is an exact example of something a vector dbs would be terrible at. Vector dbs work by fetching segments that are similar in topics to the question, so like "Where did <Character> go after <thing>" will retrieve segments with locations & the character & maybe talking about <thing> as a recent event. Your question has no similarity with the segments required in any way; & it's not the segments that are wrong it's the way they relate to the rest of the story
- toss1 3y agoGood points - LLMs are ok at finding things that exist, but they have zero ability to abstract and find what is missing (actually, probably negative; they'd likely hallucinate and fill in the gaps). Which makes me wonder if the opposite, but more laborious approach might work - request it identify all characters and plot themes, then request summaries of each. You'd have to review the summaries for holes. Lotsa work, but still maybe quicker than re-reading everything yourself?
- TeMPOraL 3y ago> LLMs are ok at finding things that exist, but they have zero ability to abstract and find what is missing (actually, probably negative; they'd likely hallucinate and fill in the gaps). I feel this is mostly a prompting issue. Specifically GPT-4 shows surprising ability to abstract to some degree and work with high-level concepts, but it seems that, quite often, you need to guide it towards the right "mode" of thinking. It's like dealing with a 4 year old kid. They may be perfectly able to do something you ask them, but will keep doing something else, until you give them specific hints, several times, in different ways.
- clbrmbr 3y agoThere are other ways of using the model other than iterative forward inference (completions). You could run the model over your novel (perhaps including a preface) and look at the posterior distribution as it scans. This may not be so meaningful at the level of the token distribution, but there may be interesting ways of “spell checking” at a semantic level level. Think a thesaurus but operating at the level of whole paragraphs.
- vidarh 3y agoFirstly, I don't at all agree that they have zero ability to abstract. Doesn't fit my experience at all. A lot of the tasks I use ChatGPT for is exactly to analyse gaps in specifications etc. And have it tell me what is missing, suggest additions or ask for clarifications. It does that just fine. But I've started experimenting with the second part, of sorts, not to find plot holes but to have it create character sheets for my series of novels for my own reference. Basically have it maintain a sheet and feed it chunks of one or more chapters and asking it to output an a new sheet augmented with the new details. With a 100K context window I might just test doing it over while novels or much larger chunks of one.
- fzliu 3y agoI'm also not entirely convinced by "huge" context models just yet, especially as it relates to fuzzy knowledge such as overarching themes or writing style. In particular, there are 0 mentions of the phrase "machine learning" in The Great Gatsby, so adding one sentence that introduces the phrase should be easy for self-attention to pick out.
- EGreg 3y agoThis sounds like all the other skepticism about what AI can do. And then it can spot 200x more than any human and correlate it into common themes, and you’ll say what?
- devmor 3y agoDoing more than a human can isn't impressive. Most computer problems for any purpose can do more of something, or something faster than a human can. A better comparison would be if it can pick out any differences that can't be picked out by more traditional and simple algorithms.
- EGreg 3y agoOr course it can very soon, since those were also written by humans. Like AlphaZero vs Rybka
- chaxor 3y agoIt does, using this method. My immediate thought as well was '... Yeah, well vimdiff can do that in milliseconds rather than 22 seconds' - but that's obviously missing the point entirely. Of course, we need to tell people to use the right tool for the job, and that will be more and more important to remind people of now. However, it's pretty clear that the reason they used this task is to give something simple to understand what was done in a very simple example. Of course it can do more semantic understanding related tasks, because that's what the model does. So, without looking at the details we all know that it can summarize full books, give thematic differences between two books, write what a book may be like if a character switch from one book to another is done, etc. If it doesn't do these things (not just badly, but can't at all) I would be surprised. If it does them, but badly, I wouldn't be surprised, but it also wouldn't be mind bending to see it do better than any human at the task as well.
- SkyPuncher 3y agoFurther, the problem with this example is it relies on a comparison against public data. Most of these AI start failing pretty hard when you ask it to do the same task on something completely novel to it (like a company document). Sometimes they'll get it right. Other times, they'll spit out gibberish that's clearly some generic answer.
- nomel 3y ago> Most of these AI This is as meaningful as saying most of the hominids can't count. You can't usefully generalize AI models with the rate of change that exists right now. Any statements/comparisons about AI has to contain specific models and versions, otherwise it's increasingly irrelevant noise.
- jiggawatts 3y agoEvery time someone has said "LLMs can't do X", I tried X in GPT 4 and it could do it. They usually try free LLMs like Bard or GPT 3 and assume that the results generalise.
- dmix 3y agoI'd imagine working with an entire company document would require a lot more hand holding and investment in prompt engineering. You can definitely get better results if you add much more context of what you're expecting and how the LLM should do it. Treating these LLMs as just simple Q&A machines is usually not enough unless you're doing simple stuff.
- coffeebeqn 3y agoWhat techniques do they actually use to achieve 100k? I assumed they load the document to a vector databases and then do some kind of views into that
- raymond_goo 3y agoAlibi: https://arxiv.org/abs/2108.12409 https://arxiv.org/abs/2108.12409 https://www.reddit.com/r/MachineLearning/comments/ww146r/d_alibi_enables_transformer_lms_to_extrapolate_to/ https://www.reddit.com/r/MachineLearning/comments/ww146r/d_a...
- asadm 3y agoI would assume by training LLM on dataset of 100k tokens would be the right way.
- clbrmbr 3y agoMakes me wonder whether we could get really huge contexts much more efficiently by feeding back a higher layer back into the tail end of the model. That way it has a very clear picture of the recent text but only a compressed picture of the earlier parts of the document. (I think I’ve got to read up on how transformers actually work.)
- desperate 3y agoAfaik you're describing something akin to a recurrent neural network, and the problem with that is that it doesn't parallelize well to modern hardware. And vanishing gradients.
- jiggawatts 3y agoI had the same thought as the comment you're responding to. Recurrent neural networks are bad when the recurrence is 100x long or more. You need long chains because with a token-at-a-time, that's what you need to process even one paragraph. But if you use an RNN around a Transformed-based LLM, then you're adding +4K or +8K tokens per recurrence, not +1. E.g.: GPT 4 32K would need just 4x RNN steps to reach 128K tokens!
- hfhdndnfbdbd 3y agoI think people in the comments are missing the point. I might be wrong, but the point isn't comparing a modified The Great Gatsby to the original one. Of course that's not impressive and it's an easy thing to do. The point of the exercise is supposed to be[1] that the model has the entire novel as context / prompt and so can identify (within that context), whether a paragraph is out of place. That is impressive and I wouldn't know how to find that programmatically (would you have a list of "modern" words to check? But maybe the out of place thing is hidden in the meaning and there's no out of place, modern word). [1] I say supposed to be because the great Gatsby is in the training sample and so maybe there is a sense in which the model "contains" the original text and in some way is doing "just" a comparison. A better test would be try with a novel or document that the model hasn't seen...or at least something not as famous as the great Gatsby.