This feels like a failure to learn the bitter lesson: You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
Large Concept Models: Language modeling in a sentence representation space
11–20 of 61 posts
Re: Large Concept Models: Language modeling in a sentence representation space
#12This feels like a failure to learn the bitter lesson: You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
That should be proven. The two approaches - predicting tokens vs predicting "sentences" - should be compared to see how much their output differ in terms of quality.
Edit2: ...and both (and their variants) be compared to other ideas such as "multi-token prediction"...
Edit: or, appropriateness of the approach should be demonstrated after acquired "transparency" of how the LLMs effectively internally work. I am not aware of studies that make the inner workings of LLMs adequately clear.
Edit3: Substantially, the architecture should be as solid as possible (and results should reflect that).
Re: Large Concept Models: Language modeling in a sentence representation space
#13This feels like a failure to learn the bitter lesson: You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
The bitter lesson isn’t a law of nature, though. And as GPT-style LLMs appear to be at the foot of a scaling wall, I personally think inductive bias is due for a comeback.
Re: Large Concept Models: Language modeling in a sentence representation space
#14This feels like a failure to learn the bitter lesson: You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
The bitter lesson isn’t a law of nature, though. And as GPT-style LLMs appear to be at the foot of a scaling wall, I personally think inductive bias is due for a comeback.
Everyone keeps claiming this but we have zero evidence of any kind of scaling wall what-so-ever. Oh you mean data? Synthetic Data, Agents, and Digitization solve that.
Re: Large Concept Models: Language modeling in a sentence representation space
#15This feels like a failure to learn the bitter lesson: You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
It is explicitly stated in the paper that
> One may argue that LLMs are implicitly learning a hierarchical representation, but we stipulate that models with an explicit hierarchical architecture are better suited to create coherent long-form output
And the problem remains that (text surrounding the above):
> Despite the undeniable success of LLMs and continued progress, all current LLMs miss a crucial characteristic of human intelligence: explicit reasoning and planning at multiple levels of abstraction. The human brain does not operate at the word level only. We usually have a top-down process to solve a complex task or compose a long document: we first plan at a higher level the overall structure, and then step-by-step, add details at lower levels of abstraction. [...] Imagine a researcher giving a fifteen-minute talk. In such a situation, researchers do not usually prepare detailed speeches by writing out every single word they will pronounce. Instead, they outline a flow of higher-level ideas they want to communicate. Should they give the same talk multiple times, the actual words being spoken may differ, the talk could even be given in different languages, but the flow of higher-level abstract ideas will remain the same. Similarly, when writing a research paper or essay on a specific topic, humans usually start by preparing an outline that structures the whole document into sections, which they then refine iteratively. Humans also detect and remember dependencies between the different parts of a longer document at an abstract level. If we expand on our previous research writing example, keeping track of dependencies means that we need to provide results for each of the experiment mentioned in the introduction. Finally, when processing and analyzing information, humans rarely consider every single word in a large document. Instead, we use a hierarchical approach: we remember which part of a long document we should search to find a specific piece of information. To the best of our knowledge, this explicit hierarchical structure of information processing and generation, at an abstract level, independent of any instantiation in a particular language or modality, cannot be found in any of the current LLMs
Re: Large Concept Models: Language modeling in a sentence representation space
#16Earlier quoted context omitted.
The bitter lesson isn’t a law of nature, though. And as GPT-style LLMs appear to be at the foot of a scaling wall, I personally think inductive bias is due for a comeback.
Everyone keeps claiming this but we have zero evidence of any kind of scaling wall what-so-ever. Oh you mean data? Synthetic Data, Agents, and Digitization solve that.
I disagree, but I also wasn’t referring to the exhaustion of training materials. I am referring to the fact that exponentially more compute is required to achieve linear gains in performance. At some point, it just won’t be feasible to do $50B training runs, you know?
Re: Large Concept Models: Language modeling in a sentence representation space
#17Earlier quoted context omitted.
The bitter lesson isn’t a law of nature, though. And as GPT-style LLMs appear to be at the foot of a scaling wall, I personally think inductive bias is due for a comeback.
Everyone keeps claiming this but we have zero evidence of any kind of scaling wall what-so-ever. Oh you mean data? Synthetic Data, Agents, and Digitization solve that.
There were multiple reports confirming that OpenAI's Orion (planned to be GPT-5) yielded unexpectedly weak results.
Re: Large Concept Models: Language modeling in a sentence representation space
#18Earlier quoted context omitted.
Everyone keeps claiming this but we have zero evidence of any kind of scaling wall what-so-ever. Oh you mean data? Synthetic Data, Agents, and Digitization solve that.
I disagree, but I also wasn’t referring to the exhaustion of training materials. I am referring to the fact that exponentially more compute is required to achieve linear gains in performance. At some point, it just won’t be feasible to do $50B training runs, you know?
50B still seems reasonable compared to the revenue of the Big AI companies.
Re: Large Concept Models: Language modeling in a sentence representation space
#19Earlier quoted context omitted.
I disagree, but I also wasn’t referring to the exhaustion of training materials. I am referring to the fact that exponentially more compute is required to achieve linear gains in performance. At some point, it just won’t be feasible to do $50B training runs, you know?
50B still seems reasonable compared to the revenue of the Big AI companies.
what revenues? If by big AI companies you mean llm service providers (OpenAI, ...), their revenues are far from high or profitable.
https://www.cnbc.com/2024/09/27/openai-sees-5-billion-loss-t...
Maybe Nvidia, but they are a chip / hardware maker first. And even for them 50B training run with no exponential gains seems unreasonable.
Better to optimize the architecture / approach first, which also is what most companies are doing now before scaling out.
Re: Large Concept Models: Language modeling in a sentence representation space
#20This feels like a failure to learn the bitter lesson: You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
> You're just taking the translation to concepts that the LLM is certainly already doing and trying to make it explicitly forced.
That is what tokens are doing in the first place though, and you get better results with tokens instead of letters.