Which one is more important: more parameters or more computation? (2021)
50 points
1 day ago
| 3 comments
| parl.ai
| HN
vorticalbox
8 hours ago
[-]
This reminds me of https://dnhkng.github.io/posts/rys/

David looks into the LLM finds the thinking layers and cut duplicates then and put them back to back.

This increases the LLM scores with basically no over head.

Very interesting read.

reply
renticulous
5 hours ago
[-]
Jeff Dean says models hallucinate because their training data is "squishy."

But what's in the context window is sharp, the exact text or video frame right in front of them.

The goal is to bring more of the world into that context.

Compression gives it intuition. Context gives it precision.

Imagine if we could extract the model's reasoning core and plug it anywhere we want.

reply
2ndorderthought
4 hours ago
[-]
LLMs "hallucinate" because they are stochastic processes predicting the next word without any guarantees at being correct or truthful. It's literally an unavoidable fact unless we change the modelling approach. Which very few people are bothering to attempt right now.

Training data quality does matter but even with "perfect" data and a prompt in the training data it can still happen. LLMs don't actually know anything and they also don't know what they don't know.

https://arxiv.org/abs/2401.11817

reply
electroglyph
3 hours ago
[-]
> they also don't know what they don't know

they sort of do tho:

https://transformer-circuits.pub/2025/introspection/index.ht...

reply
2ndorderthought
2 hours ago
[-]
I won't quibble even though I likely should. Have to remember this is HN and companies need to shill their work otherwise ... Yes.

I will play along and assume this is sound. 10-40% +/- 10% is along the lines of "sort of" in a completely unreliable, unguaranteed and unproven way sure.

reply
majormajor
1 hour ago
[-]
That just sounds like a very fancy/marketing way of saying "models will hallucinate because you cannot compress all the facts in the world into the model size." (Without even getting into any other things that could cause plausible-but-incorrect output.)

>Imagine if we could extract the model's reasoning core and plug it anywhere we want.

Aren't a lot of the latest model variants doing something very similar? Stuff more domain-relevant knowledge into the model itself on top of a core generally-good reasoning piece, to reduce need to perfectly handle giant context?

reply
kang
5 hours ago
[-]
The answer should be obvious that its both.

Zurada was one of our AI textbook that makes it visual that right from a simple classifier to a large language model, we are mathematically creating a shape(, that the signal interacts with). More parameters would mean shape can be curved in more ways and more data means the curve is getting hi-definition.

They reach something with data, treating neural network as blackbox, which could be derived mathematically using the information we know.

reply
anon373839
40 minutes ago
[-]
Well both aren’t “more important”, since that’s illogical. I think recent strides in high performance small LLMs have shown that the tasks LLMs are useful for may not require the level of representational capacity that trillion-parameter models offer.

However: the labs releasing these high-intelligence-density models are getting them by first training much larger models and then distilling down. So the most interesting question to me is, how can we accelerate learning in small networks to avoid the necessity of training huge teacher networks?

reply
mskogly
4 hours ago
[-]
Selective training data, lora fine tuning or MOE are other solutionsZ Sure, creating a model with 100 billion parameters will yield good results, but it’s sort of like employing a million random people to play darts. Or shooting sparrows with A nuclear bomb.
reply