Building things / previously eng @malbeclabs_xyz @DigitalOcean @Shopify
Joined April 2009
- Tweets2.2K
- Following1.3K
- Followers1.1K
- Likes2.7K
Steven Normore retweeted
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
Steven Normore retweeted
It's pencils down, people. Writing code by hand is no longer an economically viable skill for most programmers at most companies. But the future of making software has never been brighter. Don't you dare black pill this beautiful moment! youtu.be/vDjW_dRyKXY?si=6Fsf…
Steven Normore retweeted
Be like water
DoubleZero Edge now carries data from @HyperliquidX.
Streams are entirely uncapped, every market, every order, the full L4 book including HIP-3.
Introducing the first-class path to Hyperliquid’s order book, and the fastest commercially available feed on the market.
Steven Normore retweeted
yet another fun jev experiment... i took a particle fluid simulation, pulled out the part that does the physics (what happens when two particles collide) and let jev handle it at the micro scale. 6.3 million decisions for this one, and the only laws it's told to follow are conservation of momentum and energy, and it never once broke them.
Steven Normore retweeted
If you’re curious and love to learn, there has never been a better time to be alive.
Steven Normore retweeted
The big lesson from AI is that everything is code. A slide deck is code. Design is code. That cool promo video? Code. Excel automation? Code. The universe? Probably made of code too.
Steven Normore retweeted
Absolutely beautiful rant about AI in Linux Kernel from Linus yesterday:
I realize that some people really dislike AI, but this is an area
where I'm willing to absolutely put my foot down as the top-level
maintainer.
Linux is not one of those anti-AI projects, and if somebody has issues
with that, they can do the open-source thing and fork it.
Or just walk away.
AI is a tool, just like other tools we use. And it's clearly a useful one.
It may not have been that "clearly" even just a year ago, but it's no
longer in question today.
There are other questions around AI (like what the economy of it will
actually look like in the end), but "is it useful" is no longer one of
those questions. Anybody who doubts that clearly hasn't actually used
it.
Yes, it can also be a somewhat painful tool, both for maintainer
workloads and just from a "it keeps finding embarrassing bugs"
standpoint.
But the solution is not to put your head in the sand and sing "La La
La, I can't hear you" at the top of your voice like some people seem
to do.
The solution is to make sure those LLM tools _help_ maintainers
instead of just causing them pain. There's no question on that side.
We're not forcing anybody to use it, but I will very loudly ignore
people who try to argue against other people from using it.
And no, AI isn't perfect. But Christ, anybody who points to the
problems at AI had better be looking in the mirror and pointing at
themselves at the same time.
Because it's not like natural intelligence is always all that great either.
The kernel project has been and will continue to be about the technology.
Sure, the social angle of working on open source is important and
often a very motivating part of the project, but in the end that's a
side benefit, not the _point_ of the project.
This is *NOT* some kind of "social warrior" project, never has been,
and never will be.
In the kernel community we do open source because it results in better
technology, not because of religious reasons.
And so we make decisions primarily based on technical merit. Not fear
of new tools.
Linus
Steven Normore retweeted
turbopuffer crossed $100M run-rate in March. 19mo after $1M. Profitable & <$1M raised.
Cursor・Anthropic・Notion・Cognition・Harvey・Bridgewater・Ramp・Linear・Legora・Superhuman・Atlassian・Granola
We’d be nowhere without them. We work like hell to exceed their expectations.
Steven Normore retweeted
Attention isn't all you need - some attention is.
I've been exploring ways to reduce the compute and memory footprint of LLMs. I put it together in a repo: Calibrated Sparse Attention (CSA) is a one-time calibration pass on an existing model that figures out which keys each head in each layer actually uses, and you skip the rest. At matched LM quality I'm seeing potential for:
~10× longer context at the same memory + bandwidth budget
~10× less KV-read bandwidth per token (with a sparse kernel)
~10× fewer attention FLOPs
~10× more requests per GB of KV memory (composed with eviction)
No retraining or labels needed.
Steven Normore retweeted
Been on a quest to make models run with a smaller memory footprint, especially locally on my laptop, so I built a hybrid SSM+attention model with three memory layers:
• minimal attention KV for turns
• SSM state, MB-scale memory footprint for episodic
• LoRA from captures for long-term learning
Long context without the memory-bandwidth tax of GB KV caches, no prompt inflation. I'm curious to see how it scales to larger models.
github.com/jnormore/cognit
