Hacker News Reader

Comment by syllogism | Hacker News Reader

If you read comments about the paper it's pretty obvious people are misled. People think that the method read in Qwen weights directly to improve it. That's not what it did --- it authored a program to analyse and fix the weights.

What's confusing is they do have experiments that read raw weights, and the abstract is misleadingly written to make it sound like that's how the Qwen experiment works.

View on HN

I wouldn't say misled, I would just say they made up their mind about what it means regardless of what the article said. Maybe the title is misleading, but the original article is pretty fair.

but AI "authors programs" for everything ? Editing or looking at file usually ends up as a bunch of sed scripts running in shell rather than AI agent opening file directly, and that's pretty much the case for every operation

Not necessarily. Language models can do stuff—well, figure out or determine stuff—purely through inference. Give them the input on their context window and look at the output, no need to generate or run any programs.

You can take it further with a reasoning model and let it generate + consume a bunch of additional tokens before getting your output. Still not writing or running any programs.

To my understanding, that's how the examples of determining which game a small neutral net was trained to play. The weights of the neural net for into the context window, and then Astra was able to analyze them purely by reasoning rather than running external tools.

Did you really insert emdashes in the first sentence yourself?

Yes. I've been abusing em-dashes over a decade and like them too much to give them up now :)

They're pretty easy to type on Android keyboards and on macOS (option + shift + -). You can set up a similar compose key thing on Linux.

> it authored a program to analyse and fix the weights.

This is a difference without much distinction, at least to me. It's like being upset when someone says that Claude made a http call -- of course the LLM didn't make the call itself, it used a tool to do it.

It obviously is an important distinction to the paper's authors, which is why they used the modifier "native" to mark it as a distinct form of communication than what's already been on the market for years (coding agents). But it's not native, or rather some of the results were native but they mixed toy-scale native results with non-toy-scale non-native results for marketing purposes.

This is like someone telling you they made a "native" brain-to-brain interface but it's just a tin-can phone.

No it's a huge difference. Someone who can quickly compute 123^456 in his head has a very different skillset to someone who uses a calculator.

The distinction here is the equivalent of Astra passing the pelican test by prompting an image generation model versus emitting the SVG.

The former will likely produce a much higher quality image, but claiming it as a demonstration of Astra's capabilities would be a lie.

Reply on news.ycombinator.com