The Odyssey post credits scene went hard youtube.com/watch?v=0Vyj1C8o…
mtobi retweeted
Solstice - 5: Forgotten Archives is now available on youtube!!
youtu.be/Gl2hTmgG18k?si=PsEY…
make sure to watch it in 4K with best sound device!!
mtobi retweeted
4 years ago, I left my successful academic career to pioneer a new kind of life as an independent scholar on the internet.
Took me 3 years to hit six-figures annual income.
This Friday, I'm publishing my playbook (free).
If you follow me, REPLY or RT this and I'll DM it to you
mtobi retweeted
He’s right. 100% code coverage is useless because of complex branch conditions.
You really want 100% Modified Condition/Decision Coverage (MC/DC) — a meaningful code-coverage metric, but one not generally taught.
tc.faa.gov/its/worldpac/tech…
mtobi retweeted
super useful VSC extension for previewing models. it even has a scene inspector and generates *.jsx components on click. 😱
mtobi retweeted
I beat Teardown a while ago but this amazing frame rendering article breakdown by @unconed just got me wanting to play it again. Really cool transitions and super insightful articles his site!
acko.net/
#threejs #webgl #creativecoding
Roger Thisdell discusses the Big Mind / God Consciousness state (3/5) of spiritual development. Within his system, it is above the Witness and below No Self (see other videos).
youtu.be/HWkZI0W6q5o
mtobi retweeted
LongNet: Scaling Transformers to 1,000,000,000 Tokens
Presents LONGNET, a Transformer variant that can scale sequence length to more than 1 billion tokens, without sacrificing the performance on shorter sequences
abs: arxiv.org/abs/2307.02486
repo: github.com/microsoft/torchsc…
mtobi retweeted
Did you know that people tried to prove central limit theorem for over two centuries, first starting with de Moivre (1733), then almost a century after by Laplace who both used binomial distribution.
Then it was Poisson who worked on this theorem, and Chebyshev (1890–1891) who gave a rigorous demonstration of it in the middle of the nineteenth century.
At the beginning of the twentieth century, the Russian mathematician Liapounov, Aleksandr Mikhailovich (1901) created the generally recognized form of the central limit theorem by introducing its characteristic functions.
Markov, Andrei Andreevich(1908) also worked on it and was the first to generalize the theorem to the case of independent variables.
In 1924 Kolmogorov started to become interested in research in Probability Theory and in 1928 he was able for the first time to formulate necessary and sufficient conditions of the Law of Large Numbers that escaped other best mathematicians of the time for many decades.
It has taken the best mathematicians almost two centuries to prove conditions for LLN and prove CLT.
In fact there is almost 500 (!) pages book describing the history of CLT.
medicine.mcgill.ca/epidemiol…
#statistics #machinelearning #gaussian
mtobi retweeted
tinyvector - the tiny, least-dumb, speedy vector embedding database.
pretty much: you don't need complicated algos, just brute force nearest neighbors.
pre-launching this project + why i'm building this:
mtobi retweeted
trying lastest realism-effects by @0beqz (it's the poisson-recursive branch github.com/0beqz/realism-eff…). he's made incredible progress. it looks better than blender EEVEE imo, it casts shadows, converges quickly and there's less ghosting. waiting for the official release. 🎉
mtobi retweeted
Pipeline has correctly settled on hack pipes, which is stage 2
Pattern matching is stage 1
C’mon tc39 ship these 2 bad boys and JS will be so fucking 🔥
github.com/tc39/proposal-pip…
github.com/tc39/proposal-pat…
gave this library a try to relive some elixir glory days and the amount it cleans up my code is ridiculous: github.com/gvergnaud/ts-patt…
pattern matching and pipes need to be added to JS ASAP
mtobi retweeted
Effect of cooling of cooked white rice on resistant starch content and glycemic response
pubmed.ncbi.nlm.nih.gov/2669…
mtobi retweeted
LLMs can now fit into your phone.
Sparse-Quantized Representation (SpQR) is a brand new technique to compress LLMs.
It identifies outlier weights that cause large quantization errors and stores them at higher precision, while compressing the rest to 3-4 bits. This results in less than 1% accuracy loss, which is very impressive!
The potential of SpQR is huge, You can run a 33B LLM on a single 24GB GPU fully lossless.
arxiv.org/abs/2306.03078🧵
mtobi retweeted
Wooow the @EffectTS_ introduction almost reached 100 stars on GitHub in few days 🌟 I received a lot of positive feedbacks and ideas to improve it and that’s super great, please feel free to keep going and open issues!
github.com/antoine-coulon/ef…
mtobi retweeted
Glenn Gould's "How to write a fugue" is a fun self-documenting/commenting piece. I love "heavy meta" music and other arts that analyse and exhibit their forms from inside, using themselves as examples. youtube.com/watch?v=pHW1I8T0…





























