6/26/2026 at 10:36:47 PM
I really wish more people skeptical of AI capabilities would read about scaling laws -- Lilian is always so marvelous at giving a deep overview of the technical side but the whole point of this is: there are scaling laws, and they hold and continue to hold. This is such a huge basis for the predictions about AI capabilities for the past like 5 years.by aspenmartin
6/26/2026 at 11:42:37 PM
And sitting right next to the data and compute factors in every cross entropy loss equation is the entropy of the language, which is just a fixed constant. There’s such a hard cap on cross entropy loss training and I never hear it come up!by FromTheFirstIn
6/27/2026 at 1:03:33 AM
Right but that is context dependent; it drops with context length, depends on tokenizer, etc. It doesn't end up being super relevant, despite the fact that if you look at the loss for real models it's relatively large in absolute terms. But that doesn't really matter -- all of the interesting stuff happens once you start getting closer and closer to it. You've gotten past all of the easy tokens that dominate the entropy and now you get to the really challenging ones that we care about (like e.g. very difficult reasoning about a next step).by aspenmartin
6/27/2026 at 1:20:15 AM
My understanding is that the true entropy floor of a language is intractable- regardless of context length there will be “unpredictable” tokens where cross entropy loss is bound to happen. Even with infinite parameters and data you’ll still have a chance at failing to predict the next token correctly a decent chunk of the time.Also, linear gains in context length scale quadratically with compute because of attention, so depending on context growth means taking a bath on GPUs for as long as you can, right?
by FromTheFirstIn
6/27/2026 at 7:11:01 AM
Yeah I mean, if you and I were to play the word-guessing game where you needed to guess what next word I'm thinking of, there's always uncertainty in your guess because it's a game of partial information - you can't fully observe my inner state. But that doesn't mean you couldn't evolve a strategy that spends a really long time thinking and analyzing to get asymptotically close to the best guess. There's no limit on that intelligence.by graboy
6/27/2026 at 12:13:16 PM
Isn’t the limit exactly what you’re describing? There’s always uncertainty, and your asymptote can approach its limit but it does have a limit. That’s the limit to the intelligence. And this is just for cross entropy loss- even if you could get loss to 0, I’m still not convinced at all that an enormous semantic map and its convoluted geometries amounts to intelligence.by FromTheFirstIn
6/27/2026 at 3:58:03 PM
If you get to E you have generated a Bayes-optimal model of the conditional distribution (as in, next token conditional on context). This is something I thought too, but even if you're a fraction of a nat above the floor, you could have enormous headroom in performance left because there are still rare tokens amongst the irreducible noise that require so much capability to predict. It's not to suggest there truly is no cap on capability, but just that this constant isn't really saying what that is.by aspenmartin
6/27/2026 at 4:17:56 PM
Yeah, it not a linear cap (x% entropy doesn’t mean x% wrong) but it does seem like a hard cap. To be honest, the more I’ve understood scaling laws the more I think that the elephant in the LLM room is the entropy of the language. It explains why coding languages are so much more tractable (they’ve got WAY less entropy) and it explains why we haven’t seen a step function in capabilities for LLMs since GPT-4 outside of making specific toolings for particular contexts. I think E is coming to dominate and there isn’t a workaround for it.by FromTheFirstIn