Qwen 3.8 27B on a laptop can read raw model weights and guess the game
Qwen 3.8 27B on my laptop can interpret model weights, too
Simon Willison tested a paper claiming LLMs can interpret neural network weights without running code. Feeding raw floating-point weights from four small game-playing networks to Qwen 3.8 27B on his laptop, the model correctly identified Pac-Man, Pong, and Breakout — taking 46 minutes to over an hour each — but missed the more obscure Freeway. He pushes back on claims that source code is obsolete.
I'm not ready to declare that the age of source code is over and our future involves incomprehensible machines editing their own model weights in incomprehensible ways.
- variaga
Not making any claims about a specific model's capabilities, but all these "can interpret model weights" posts brings to mind the line in Shakespeare's Henry IV
GLENDOWE: I can call spirits from the vasty deep.
HOTSPUR: Why, so can I, or so can any man, but will they come when you do call for them?
I can interpret weights too, but will that interpretation be correct?
- zkmon
Qwen-3.8-27B - the best thing that happened in AI so far.
This thing can accomplish the same results as the frontier models do, if you are willing to play a role of an higher level architect or decision maker. Guide it a bit here and there like a tech lead and you get everything done.
The money you pay for frontier models is exactly for outsourcing that high level direction work to agents.
- athrowaway3z
Hopefully we get a new benchmark: How well can a model tweak a set of weights to produce pelican on a bike?
- data-ottawa
Perhaps I’m missing this, but can Qwen 3.8 alter Qwen 3.8 weights?
There are a lot of known neurons in the older Qwen models that are long known to induce behaviour changes (see evil Qwen by changing one weight).
My question boils down to how much is trained vs being able to actually understand the weights. Presumably a same gen model isn’t trained on any sources dissecting that generation. And how much do longer known conceits like steering vectors, J spaces, LoRAs play into this?
This doesn’t explain the game training cases, which are very impressive to me.
- layer8
> I’m not ready to declare that the age of source code is over and our future involves incomprehensible machines editing their own model weights in incomprehensible ways.
That sounds a bit like biology; maybe AI will become like that.
- olalonde
I was thinking along these lines a couple of weeks ago[0] - specifically whether we could treat weight manipulation as a sort of programming/compilation task. I expected it to require a lot more training with synthetic datasets. It's surprising that existing models are already somewhat decent at it.
- NitpickLawyer
Simon, could you try reproducing Experiment 5 from the blog post, if you have a chance? Apparently this person tried [1] with API models and only the latest models can do the impressive 10-20x vs. PPO baseline.
[1] - https://nitter.netbub.com/AlexKrentsel/status/21173766837246...
- mirekrusin
So who's first going to make model look, update and reload its weights in a loop?