logoalt Hacker News

Toutouxc • today at 6:05 PM • 4 replies • view on HN

I’m hoping someone can explain this to me. They say “no training”:

> for the first time, a language model rewrites its own weights by its own reasoning and becomes better, with no gradients and no training

> it tests its edits on practice problems and submits only what works, each edit improving on the last, and over eight consecutive sessions its accuracy only rises

But this sounds a lot like training, and “submitting only what works”, while probably not gradient descent by definition, also seems kind of obvious when you’re optimizing something.


Replies

fl1pper • today at 6:33 PM

The paper states that this approach doesn't use backpropagation, an optimizer, or a parameter-fitting loop.

What makes it work is that it reads its own hidden states and weight matrices -> then writes a few-line Python script to fix the weights directly (within a 20-forward-pass budget, so it can't just brute-force guesses).

After skimming through the paper and comparing it with how SEAL works (where the model still relies on RL and gradient descent to apply its self-edits), I understood it something like that :)

But I think there are comments in the thread that explain the main concept far better than I do.

Tuna-Fish • today at 6:07 PM

What it means is that they are not doing backprop.

➕ show 1 reply
William_BB • today at 6:25 PM

Is it actually converging to something meaningful? You could write the same procedure in a simple deterministic python script, that doesn't mean it's not learning garbage in the long run.

anon291 • today at 6:26 PM

Training means we take the mathematical expression corresponding to the entire model, use differentiation rules to generate partial derivatives according to a user-specified loss function (not part of inference), and then apply the gradients from the calculated differentials using a learning rate.

In this case, they just ask the model what the weights ought to be.