parquet packfiles (or parqfiles as I like to call them) are 17% smaller, but packfiles are generally faster to work with
this is storing non deltified objects in parquet and relying on zstd (level 9) for compression
(ps: implementation is vibe coded, I don't have time)
69
alex retweeted
Sometimes on campus you accidentally walk by a Comp Sci class and the professor is writing like
SELECT title, genre, price
FROM books
WHERE (genre = 'Fantasy' OR genre = 'Sci-Fi') AND price < 12.00;
on the board and everyone is taking notes like its actual school
alex 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.…
alex retweeted
1/3
A CPU inside an LLM is one thing, but running another LLM inside an LLM would be taking things a bit too far, wouldn't it?
Anyways, here's an LLM running inside an LLM : QwenPU, an LLM (@karpathy's stories260k) running on an NPU compiled inside Qwen 3.5-4B's weights.















