LessWrong 2.0 Reader

Why does California have forests so close to residential areas if they can not handle wildfires? In a just world, insurance companies would be allowed to pave over Californian forests with concrete.

The alternative hypothesis does need to be said, especially after someone at a party outright claimed it was obviously true, and with the general consensus that the previous export controls were not all that tight. That alternative hypothesis is that DeepSeek is lying and actually used a lot more compute and chips it isn’t supposed to have. I can’t rule it out.

Re DeepSeek cost-efficiency, we are seeing more claims pointing in that direction.

In a similarly unverified claim, the founder of 01.ai (who is sufficiently known in the US according to https://en.wikipedia.org/wiki/Kai-Fu_Lee) seems to be claiming that the training cost of their Yi-Lightning model is only 3 million dollars or so. Yi-Lightning is a very strong model released in mid-Oct-2024 (when one compares it to DeepSeek-V3, one might want to check "math" and "coding" subcategories on https://lmarena.ai/?leaderboard; the sources for the cost claim are https://x.com/tsarnick/status/1856446610974355632 and https://www.tomshardware.com/tech-industry/artificial-intelligence/chinese-company-trained-gpt-4-rival-with-just-2-000-gpus-01-ai-spent-usd3m-compared-to-openais-usd80m-to-usd100m, and we probably should similarly take this with a grain of salt).

But all this does seem to be well within what's possible. Here is the famous https://github.com/KellerJordan/modded-nanogpt ongoing competition, and it took people about 8 months to accelerate Andrej Karpathy's PyTorch GPT-2 trainer from llm.c by 14x on a 124M parameter GPT-2 (what's even more remarkable is that almost all that acceleration is due to better sample efficiency with the required training data dropping from 10 billion tokens to 0.73 billion tokens on the same training set with the fixed order of training tokens).

Some of the techniques used by the community pursuing this might not scale to really large models, but most of them probably would scale (as we see in their mid-Oct experiment demonstrating scaling of what has been 3-4x acceleration back then to the 1.5B version).

So when an org is claiming 10x-20x efficiency jump compared to what it presumably took a year or more ago, I am inclined to say, "why not, and probably the leaders are also in possession of similar techniques now, even if they are less pressed by compute shortage".

The real question is how fast will these numbers continue to go down for the similar levels of performance... It's has been very expensive to be the very first org achieving a given new level, but the cost seems to be dropping rapidly for the followers...

Great post! Extremely interested in how this turns out, I’ve also found:

Alignment faking could be just one of many ways LLMs can fulfill conflicting goals using motivated reasoning.

One thing that I’ve noticed is that models are very good at justifying behavior in terms of following previously held goals.

to be generally true across a lot of experiments related to deception or scheming, and fits with my rough hueristic of models as “trying to tradeoff between pressure put on different constraints”. I’d predict that some variant of Experiment 2 for example would work.

Yeah. I definitely find myself mostly in the “run once” regime. Though for stuff I reuse a lot, I usually invest the effort to make nice frameworks / libraries

I mean, I think you are right about him being anti-progress and fantasy-proposing terrible 'solutions' to real problems. I don't think you are correct to give him so much credit for productivity slowdown. I think his effect is quite a bit more minor than that. I think the productivity slowdown is mostly due to weird unanticipated long term downstream effects of our government structure, land use rules, tax structures, and a tendency towards veto-ochracy.

That's exactly o1 full is not available in API as well. Why not make money on something you already have?

The LLMs are presented with the ML task and they write python code to solve the ML task. This python code is what is run in the isolated docker with 12GB memory.

So the LLMs themselves are not run on the TITAN V, they are mostly called through an API. Although I did in fact run a bunch of the LLMs locally through ollama, just not on the TITAN V server, but a larger one.

I think you are right that most people suffer from this lack of big ambition. I think I tend to fail in the other direction. I consistently come up with big plans, which might potentially have big payoffs if I could manage them, and then fail partway through. I'm a lot more effective (on average) working in a team where my big ideas get curtailed and my focus is kept on the achievable. I do also think that I bring value to under-aimers, by encouraging them to think bigger.

Sometimes people literally laugh out loud at me when I tell them about my current goals. I am not poorly calibrated, overall. I tell them that I know my chance of succeeding at my current goal is small, but that the payoff would really matter if I did manage it. In the course of aiming at a long term goal, I do tend to actively try to set aside my realistic estimate of my success, in order to let myself be buoyed by a sense of impending achievement. Being too realistic in the 'doing' phase, rather than the 'planning' phase tends to sharply bring down my chance of sticking with the project long enough that it has a chance to succeed.

Hi. A question here.

"The system executes code in a Docker container with strict resource limits (TITAN V GPU with 12GB memory, 600-second timeout). This ensures fair comparison between models and tests their ability to work within realistic constraints."

How can you run llama-3.1/3.3-70b models with 12GB vram?

Tagging @gwern [LW · GW] here for your take on the situation.