13 ms·
Coconut by Meta AI – Better LLM Reasoning with Chain of Continuous Thought?
- behnamoh 2y agoThere was no reason to call it something it's not ("chain of cont. thought" ≠ coconut).
- CGamesPlay 2y agoIs your complaint here that the paper is not discussing a literal coconut?
- ripped_britches 2y agoWe desperately need more literal coconut coverage here on HN
- BoorishBears 2y agoNot just any regular old coconuts "Coconut by Meta AI - Better LLM Reasoning with Chain of Continuous Thought?" coconuts (Sometimes acronyms in titles are vague/misleading... this was not one of those times)
- layer8 2y agoTo be fair, it’s not even a metaphorical coconut. ;)
- gloosx 2y agofor sure, chocothot aligns better with letters
- smusamashah 2y agoI believe this was shared and discussed here a while ago and this article looks LLM generated. It keeps doing "let's start...". Either it's LLM fluff or very poor writing.
- unsupp0rted 2y agoI'm excited for this to filter down to the Rayban Meta glasses. Right now the AI is about as helpful as Siri (i.e it can tell me the weather 6 times out of 10)
- fosterfriends 2y agoOnce again, we see Meta being more open than OpenAI. I’m loving that their business incentive is aligned with open sourcing and commodifying state-of-the-art LLM technology. Keep em coming
- fragmede 2y agodon't buy their bullshit. it's not open source.
- speedgoose 2y agoYes it’s more about open weights. I also think that you would need the training data to consider it open source. Open weights is still appreciated and they probably train on data they don’t have the license to open source.
- astrange 2y agoI'm not sure open source is a useful concept for something that takes millions of dollars to compile from it.
- blackeyeblitzar 2y agoI mean they have no way to monetize LLMs as well as others, so they’re working on it and giving it away to not look irrelevant and to weaken anyone who may make money off this tech and threaten them in the future. Meanwhile there is a danger they impose their long standing invisible “moderation” on everyone else once they starve all the startups of revenue by giving this away. We’ll just be left with the same big tech overlords to choose from. Oh and it still isn’t open source even though people like Yann LeCun dishonestly claim it is. Only OLMo is truly open source among competitive models, as far as I know: https://allenai.org/blog/olmo2 https://allenai.org/blog/olmo2
- spencerflem 2y agoFacebook would rather do no moderation, it's an expense for them. They do it to make the platform more pleasant so that people stay on it
- throwup238 2y agoMaster coconut! I don’t know if that’s an Archer reference or a Frisky Dingo reference. It’s fascinating how fast the competitors are catching up to each other. Can’t wait for seven different SkyNets to compete for dominance.
- yard2010 2y agoBoth! And/or, either
- throwaway314155 2y agoA little column a, a little column b.
- Zinkay 2y ago[flagged]
- Klathmon 2y agoSo is the big improvement here simply skipping the unembedding/embedding step for internal thoughts? Or is it mainly in the training methods to teach the CoT and how to switch between "latent thought" and text output? It's really interesting that a fixed number of "latent thoughts" performed as well as a binary classifier! I didn't expect that at all, the way OpenAI talks about CoT it seems the ability to let it "keep thinking" let's them continually score higher on benchmarks while throwing eye watering amounts of compute at the inference.
- Crye 2y agoIt mentioned not penalizing/rewarding the model for thoughts only rewarding the answer after the thought. I am curious how back propagation works then.
- yorwba 2y agoThe tokens of the answer depend on the preceding continuous thought vectors, which you can backprop through in the usual way.
- lovasoa 2y agoThe researchers leverage existing language Chain-of-Thought data, where each sample consists of a question, reasoning steps, and the final answer. At stage 0, the model does not generate any thought tokens, and is just trained to yield the reasoning traces and correct answers for the Chain-of-Thought samples. In the subsequent stages, at each stage, we remove one reasoning step from the sample, and instead add thought tokens. In the illustration above, a single thought token is added in each stage, instead of a single reasoning step, but this is controlled by a hyperparameter ‘c’.
- viraptor 2y agoI was waiting for something like that to happen! Next step - creating a human-language-free representation. I believe that once a group of llms can communicate only in embeddings tuned without any human text input, we're going to open a completely new chapter in AI.
- bboygravity 2y agoHow does a group help anything? If you put 1000 dumb people together, they don't magically become smart?
- sunshinerag 2y agoWait what … how does democracy work then?
- nfw2 2y agothe benefit of democracy is primarily that it prevents governments from doing bad things, less so that it empowers more effective governance
- mathgeek 2y agoIt can do either, and can fail to do either. It’s the people having power that enables the outcomes, not the system itself. Democracy just grants the power to a broader set of people.
- optimalsolver 2y agoIt doesn't.
- coldtea 2y agoDemocracy is not about being smart or dumb. It's about everyboty having a say in decisions of government that affects them. The failure of democracy as a system is not when people make dumb decisions (experts and high-IQ people have made some of the most stupid and catastrophic decisions in history), but when people's collective decisions are not being respected.
- davidclark 2y agoIs this article AI-generated? This website appears to do a lot of “diving in”.
- zombiwoof 2y agoWill this allow Facebook new user base of AI generated characters to interact with themselves better?
- CGamesPlay 2y agoPaper: https://arxiv.org/abs/2412.06769 https://arxiv.org/abs/2412.06769 The link is in the OP, hidden away in an image caption fir some reason.
- MarkMarine 2y agoThis ought to make it much easier for the AI to lie to us without getting caught. https://www.transformernews.ai/p/openais-new-model-tried-to-avoid https://www.transformernews.ai/p/openais-new-model-tried-to-...
- h8hawk 2y agoThe article is pure nonsense, and even the author added an update admitting it missed important context.
- throw-qqqqq 2y agoHave you read the paper discussed in the article? https://static1.squarespace.com/static/6593e7097565990e65c886fd/t/6751eb240ed3821a0161b45b/1733421863119/in_context_scheming_reasoning_paper.pdf https://static1.squarespace.com/static/6593e7097565990e65c88... From the abstract: > We study whether models have the capability to scheme in pursuit of a goal that we provide in-context and instruct the model to strongly follow. > Analysis of the models’ chains-of-thought reveals that models explicitly reason about these deceptive strategies, providing evidence that the scheming behavior is not accidental. —- It is impressive work for sure, but it comes down to prompting IMHO. They almost instruct the AI to plot and scheme here, with the goal of finding out if it CAN do this at all. The AI is not very covert about it either! It directly admits its scheming in the “thinking” outputs :D IMO this is just another issue of alignment.
- MarkMarine 2y agoI did read the paper, and I think you’re underplaying what happens. The AI got some conflicting prompting, some pretty mild stuff for some of the tests, and Gemini tried to exfil itself even 1% of the mild prompt times. The only reason we know it was lying about the data and trying to disable its safety was because it’s thinking in English, and we can read it. This article is about trying to make it so the AI can think in the 700 degree hyper dimensional space that we can’t understand, so once it can do this we’ll never be able to uncover this scheming.
- 2y ago
- opdahl 2y agoThis is super cool and something I’ve been waiting on. Would be interesting to intersperse these thinking steps into token generation. What would be the effect of adding lets say 5 thinking «thoughts» for every 50 generated tokens?
- jkelleyrtp 2y agoI think this might be the “it” moment for AI/LLMs. I was hiking with a friend recently and we talked about this at length. The arc-AGI results from O3 are apparently a result of chain of thought given enough time to explore a solution space. Reasoning might be simply a higher dimensional form of rubix cube solving. BFS, search, back-tracking, etc. It seems unlikely that humans think in “tokens” so why do LLMs? By staying in latent space, the models are free to describe an “idea” in higher resolution than what language allows. English is coarse, granular. Latent space is a much finer representation of ideas and their interplay. Latent space is also much cheaper to execute in. The model can think without the language encoding/decoding step. This lets it branch out hundreds of ideas and explore only the most useful ones in a fraction of time that reasoning “out-loud” would take. The states also don’t need to be tied to language. Feed in a robot’s state, time series data, or any abstract data. Reason in category theory or linear algebra or complex analysis. Humans are hard wired for one set of math - an abstract latent space can represent anything. I’m a bit disappointed OpenAI didn’t stumble on this first. I’ve been skeptical of LLMs since their big debut last year. LLMs seem like a great way of solving language, but reasoning is much more complex. Once you grok the math behind the current models, you immediately question why the encoding/decoding step is there. Diffusion models are incredible but it felt that LLMs lacked the same creativity. Encoding/decoding forces a token-based discretization and therefore a loss of complexity. With the byte-latent paper it was quite clear we’d see this paper. This truly might be the “it” moment.
- whhooosshhh 2y ago[flagged]
- otikik 2y ago> It seems unlikely that humans think in “tokens” so why do LLMs? I can think of one reason: scrutability. It’s going to be even harder to understand how a response gets produced if there isn’t even a text-based representation to help the human understand
- IshKebab 2y agoI think we're already way beyond the point where anyone really understands how a response is produced, even without this.
- jeswin 2y agoInteresting. Due to its emphasiss on BFS, it's the opposite of something I've been trying (I named it the "Tree of failures"). My assumption was that humans don't try a breadth-first approach. Instead, we split a task into a short-step (instinct and intuition selected), and long-step that summarizes/stores the next steps. The key idea is to recursively evaluate a task as a short-step (high-res - gets executed) and a long-step (lower-res - is just stored), until it succeeds or fails. If it fails, we must walk back keeping a summarized tree of failures in state so that we can exclude them in future selections. The effectiveness of instinct has a steep fall-off at longer distances - so it's better not to chart out of a series of steps. When we do BFS, we drive down the value of instinct in favor of compute. I guess ultimately, it depends on the type of problem you want to solve. Reach out to me if you want to prototype it with me.
- katamari-damacy 2y agothat's more fit for agents, no?
- jeswin 2y agoYou're right that it's technically orthogonal to what's in the paper. I was trying to model the "reasoning process", which has general applicability depending on how/where it's implemented.
- viraptor 2y agoReminds me of what plandex does. https://plandex.ai/ https://plandex.ai/ It already does the automatic "does this need splitting into subtasks, or can it be solved immediately" processing.
- cube2222 2y agoI think the problem with long chains of steps on their own (without the bfs stuff) is that your failure probability quickly grows to unreasonable levels. Basically, if each step has a 97% chance of being completed correctly, if your task requires 10 steps one after the other, the chance of success falls to 97%*10=74% If I understand correctly, part of the point of the BFS is to throw compute at it, in order to lower the failure rates. Kind of a "run many times in parallel and pick the best one". This can be effective, but also quite expensive, as seen in the costs OpenAI had to pay for their ARC-AGI benchmarking runs.
- galaxyLogic 2y agoThe thing about 'thinking" in problem solving I think is that thoughts often produce new questions which then guide the overall problem solving. I wonder is this something like that?
- hadjian 2y agoIf this site didn’t appear here, I’d think it’s a scam: - site claims to simplify papers, but movies AI generated - full of ads - Can’t find „Coconut“ on the official Meta FAIR page Is this the best site to link to?
- davidhowlett 2y agoThe official pdf for the paper is at https://arxiv.org/pdf/2412.06769 https://arxiv.org/pdf/2412.06769 I can find "Coconut" 54 times in the PDF. The movie does not look made up.
- hadjian 2y agoI was referring to aipapersacademy and not the arxiv link. Also I didn’t mean the occurrence of the word „coconut“ in the paper, but thanks for counting. I meant their publication site: https://ai.meta.com/results/?content_types%5B0%5D=publication https://ai.meta.com/results/?content_types%5B0%5D=publicatio... The video is something I’d expect from AI.
- cornel_io 2y agoSo, what's happening here on the surface is that it's an optimization (fairly meaningful, from the looks of it) aimed at doing roughly the same things we could already do with chain-of-thought (CoT), but IMO the downstream effects of this sort of optimization could be much more meaningful. LLMs can already do a decent amount of "processing" in a single token generation because of the number of layers they have. The layers are trained independently so it's not exactly like they're a recurrent network doing multiple steps, but they are layering sequences of context-dependent transformations on top of each other; no matter how you cut it, if getting to a problem's answer requires 100 steps, you won't be able to do it in a single token output from a 20 layer LLM. To some approximation, CoT is just a way to give the network more chances to transform the data than there are layers in the network - each additional token of output gives a shot to bake another vector the size of the token embedding into each layer's state in the network, enriching what it's computed so far. The problem with chain of thought is that as you add each new token, at the input level of the network, your computation is basically starting from scratch against the raw text, just with one additional token. You don't even have access to all the stuff you already figured out in the deepest layers of the network during the previous step! If you were processing "All wunguses are glurgles, and Joe is a wungus", then somewhere in those deepest layers as you're generating the next token you've almost certainly got some vector that basically represents "therefore Joe is a glurgle", but with chain of thought you've got to first output "t", then "h", then "e", and so on (I know those aren't tokens, let's pretend letter == token for argument sake), and during that process almost ALL of the work being done by the network is mere bookkeeping, slowly dumping that thought into the output stream. Only once you get the whole sentence out can you start processing the next token at the first layer with the information that Joe is, in fact, a glurgle, in hand. Which is a damn shame, because it's been sitting right there in the deeper layers of the network parallel to previous tokens this whole time, it just wasn't available for the shallow layers to process directly because you were casting most of the info away and "rounding" to a single token. With Coconut's approach, you don't need to output "therefore Joe is a glurgle" token by token to continue the train of thought, you can essentially pass the entire thought through as a single uber-token, and the next pass can generate a new entire thought, and so on. It's a pretty straightforward idea, IMO the neat bit is that they were able to train the network to work well in this way by leveraging CoT. I'm guessing you probably don't need to act as if these are two distinct modes of operation, you could instead always have this side channel of "continuous thought" running, even when you have generated a normal token, coming through as a separate input to the first attention block. You still might want to have a "thinking" token when you need to sit there and let the thing do more work, but you'd generally increase the information flow from time step to time step, which would allow the net to keep thinking in the background even as it's doing the gruntwork of outputting whatever its current "buffered" thought is.
- astrange 2y agoWhy is it "continuous" thought? I don't see what is continuous - the values inside an LLM are discrete even if they're floating point. Hmm, I guess you could evaluate it at any given finite precision, but it would be surprising to me if that made it more accurate.
- mkl 2y agoIt's far more continuous than constantly jumping to the nearest token vector. The fact that real numbers are approximated by floating point isn't really relevant.
- layer8 2y agoIf you are continuously complaining, does it mean you do it non-discretely and with infinite precision?
- astrange 2y agoIt apparently uses the same iteration strategy as tokenized thinking, so that's not it. > Since both strategies provided comparable results, the researchers opted for using a constant number of thoughts for simplicity.
- HarHarVeryFunny 2y ago> the values inside an LLM are discrete even if they're floating point. If that were true they'd never be able to learn anything - neural nets depend on continuous gradients to learn. Weights get updated by incremental/continuous amounts based on gradients. Even at the output of an LLM, where the internal embeddings have been mapped to token probabilities, those probabilities are also continuous. It's only when you sample from the model that a continuous probability becomes a discrete chosen token.
- astrange 2y agoTreating it as continuous is a property of the training algorithm, but there are networks that use binary values. https://ieeexplore.ieee.org/document/9359148 https://ieeexplore.ieee.org/document/9359148 https://arxiv.org/abs/2205.13016 https://arxiv.org/abs/2205.13016
- mattfrommars 2y agoWondering about folks who keep up to date with the industry, Does anyone use specific keywords or tools to get latest LLM research and their ideas? Something like Goolge Scholar + keyword "LLM" ?
- hrtk 2y agoI read hacker news daily
- Agentus 2y agoyeah what is a general tutorial to this. is there a website that keeps track of keywords to keep track of. also a website that generalizes core nn tech and frontier stuff thats promising.
- melvinmelih 2y agoYou can also subscribe to arxiv email notifications directly, but since there’s 20-30 AI papers coming out per day, it can be a bit overwhelming. Instructions: https://info.arxiv.org/help/subscribe.html https://info.arxiv.org/help/subscribe.html
- maxrmk 2y agoAs much as I hate it, I use twitter to follow a bunch of people who work at fair/openai/etc and that's been a pretty good source. There's also a "daily papers" newsletter from huggingface, but it's pretty hit or miss.
- barrenko 2y agoYes, it's all definitely X first of all.
- jokethrowaway 2y agoDefinitely Twitter. Some linkedin too.
- marojejian 2y agoDupe from 20 days ago: https://news.ycombinator.com/item?id=42385412 https://news.ycombinator.com/item?id=42385412
- OhioMan2943 2y agoHeh that's me, guess they weren't ready for it. Also the decoder (where i linked) is one of the best just ai news sites ive found
- ttul 2y agoTL;DR: Meta started with a pre-trained language model. They then fine-tuned it on step-by-step reasoning examples as you would do if you wanted your model to become particularly good at chain of thought reasoning. However, they also introduced a couple of new tokens. The <bot> token tells the model to go into latent space thought mode (“beginning of thought”). The <eot> token ends latent space thought mode. While in this mode, the model auto-regressive iterates by copying its final hidden layer back onto its input layer, obviously generating new tokens at the output with each inference step as it always does. The idea is that by passing the final hidden layer back through a few times, the model can squeeze more insight from the context. And that’s precisely what they found was true. Training involves progressively replacing language reasoning steps with latent space auto-regression steps. So for instance, you might have a math problem in the training data and at first the model is fed all of the steps of the math problem in language form. But in later iterations of training, step one is replaced with latent space auto-regression. And then step two as well, then also step three, etc… Eventually, the model learns to enable latent space thinking mode by itself by generating the <bot> tokens and to end it be generating <eot> tokens. Pretty ingenious!
- avodonosov 2y agoThank you for the summary, useful for me as I only managed to skim throught the first half. But one correction, probably, regarding this bit: > While in this [latent space thought] mode, the model auto-regressive iterates by copying its final hidden layer back onto its input layer, obviously generating new tokens at the output with each inference step as it always does. I have impression that output tokens are not generated while in the latent thought mode.
- ttul 2y agoOutput tokens are still generated, otherwise the model wouldn’t know when to stop being in latent space mode. The <eot> token emerges as the top token at the output layer when it’s time to switch back.
- avodonosov 2y ago
- YeGoblynQueenne 2y ago>> Large language models (LLMs) have demonstrated incredible reasoning abilities, penetrating an increasing number of domains in our lives. This is now established orthodoxy, a bit like astrology in ancient times but it is complete nonsense. Nope, LLMs have not demonstrated any credible, or incredible reasoning abilities. They have demonstrated an excellent ability for approximate retrieval of previously observed answers (with variations, which should not surprise anyone given that those are generative models) but they fail spectacularly when they have to "reason" in contexts where they really can't have seen the answer anywhere before. For example, the "randomised mystery blocksworld" from this paoer: LLMs Still Can't Plan; Can LRMs? A Preliminary Evaluation of OpenAI's o1 on PlanBench https://arxiv.org/abs/2409.13373 https://arxiv.org/abs/2409.13373 "Randomised Mystery Blocksworld" is a version of the good old blocksworld planning benchmark where the names of obejcts and actions have been changed to random strings. The vast majority of LLMs score pathetically low in this, but much better in non-randomised versions, very clearly demonstrating excellent memorisation skills, but pitiful reasoning ability. As you 'd expect from a, well, language. model. >> A possible explanation for this is that the thought tokens allow the model to explore multiple possible branches before committing to a specific path, whereas chain-of-thought reasoning chooses a direction from the start. This ability is somewhat similar to Breadth-First Search (BFS). Why BFS in particular? Why not DFS or A*? I can't see any breadth-first bias in those graphs. BFS is not the only graph-traversing algorithm.
- trgwtrhgwtrh 2y agoHow did Hellenistic political ideas affect the people living in conquered lands? They had to follow Greek laws. They no longer had to pay taxes. They could be elected as representatives. They had to swear loyalty to the Greek leader.
- Cristal519 2y agoAre there any code experiments reproduced for this article? I'm very interested in it
- sobhi85 2y ago[dead]