For the tiny models with a few thousand weights, it seems plausible Astra can actually understand what the network detects and figure out it's e.g. a triangle detector. This is possibly helped by the existence of many breakout playing networks on the web.

For Qwen, even the smallest versions like 3.5-0.8B it's becoming less plausible that it could be interpreted in a forward pass, even with reasoning. I'm wondering if the result here is recalled from many examples of instruction tuned versions of this model.

A smaller Gemma model doesn't fare as well but unclear whether this is due to fewer users and less weight examples or because it's already very well trained.

With due credit to the author and the community pushing these research fronts, I venture to say that the core capability of finding patterns in text and media comes to LLMs from the ability to find patterns in numbers in the first place. Since model weights are numbers, finding patterns and issues with those numbers should be a natural consequence, I believe.

Not so long ago everyone was amazed that they could read BASE64.

This is giving singularity is near vibes.

Seeing that they were working on Qwen3.5-0.8B has given me new hope. I would really like to get Jeff-Code working the way it was advertised, but so far, all of my results with Jeff-Code have shown it to underperform compared to direct Qwen3.8-27B. It's possible there is something wrong with the underlying Jeff v1.3 model, which is based on the same Qwen3.5-0.5B. Perhaps Astra will solve my problem.

[1] https://github.com/firelex/jeff-code

[2] https://github.com/firelex/jeff#benchmarks

Incredible results. How can we know that the modifications were not existing on the internet?

"Now let's see if Astra can write a neural network from scratch in zero-shot."

This strongly suggests retrieving pre-trained weights from memory, no?

This is wild. Could Astra read and potentially audit a mapped human brain in that case?

Absolutely insane.

If anyone's curious, the handwritten Go bot is very weak, about the level of a human complete beginner. An MLP is a very very bad architecture for a Go network, it's amazing that it does anything at all, but it's one of the least impressive results in the paper.

Play it here, "best move" is a bit stronger than "T=1", or browse games against my own tiny nets: https://claude.ai/artifact/AFyWD39J5sN2RfRe7VbpUx

On the one hand, the paper (and especially the headline) overstates and overhypes the results.

On the other hand, the overstated results, if they were true, would be tantamount to saying “we're speedrunning exactly the kind of recursive self-improvement that is a major ingredient in every human extinction scenario”, and yet it is both presented and received as an exciting opportunity.

I think it would be worth a human attempting this task....

On the face of it, running a model to a specific state, looking where is makes a mistake, and then figuring out which weights need to be jacked up to get the desired outcome doesn't seem like an impossible task...

Is it aware of projects like [1] and implementing those?

[1] https://github.com/lyramakesmusic/neuron-explorer