@Geovanny_092 Profile picture

Geo.vanny

@Geovanny_092

The fruit of the tree, belongs to the tree.

Geo.vanny Reposted

Homebrew Systems Club in Munich yesterday was a blast. So many interesting conversations and, man, what a lovely group of people. Great day. I talked about Writing Tucan, about the compiler I built in my spare time for 3 years. Here are the slides: thorstenball.com/talks/writing_…

Tweet Image 1

Geo.vanny Reposted

Finally released Hermes 3! A 405b, 70b, and 8b version! You can find all the details and resources about Hermes 3 on our website: nousresearch.com/hermes3/ We have a H3 405B bot running in our discord, join and try it out here: discord.gg/NousResearch Or give it a try on…

Tweet Image 1

Introducing 𝐇𝐞𝐫𝐦𝐞𝐬 𝟑: The latest version in our Hermes series, a generalist language model 𝐚𝐥𝐢𝐠𝐧𝐞𝐝 𝐭𝐨 𝐲𝐨𝐮. nousresearch.com/hermes3/ Hermes 3 is available in 3 sizes, 8, 70, and 405B parameters. Hermes has improvements across the board, but with particular…

Tweet Image 1


I’m done, why I even laugh this bad at this

Tweet Image 1

Geo.vanny Reposted

We often think of an "equilibrium" as something standing still, like a scale in perfect balance. But many equilibria are dynamic, like a flowing river which is never changing—yet never standing still. These dynamic equilibria are nicely described by so-called "detailed balance"


Geo.vanny Reposted

Nothing beats ol'reliable

Tweet Image 1

So I have to get better at English and put 3 pages long prompt to actually see good results?


Geo.vanny Reposted

Left is a typical kindergarten writing sample. Right is typical of a 5.5 year old who does most of her writing in Google Docs.

Tweet Image 1
Tweet Image 2

Geo.vanny Reposted

There’s a guy on tiktok who owns a company where you can complain about your boss anonymously and he goes to your work and cuss them out for you 😂


Geo.vanny Reposted

Moravec's paradox in LLM evals I was reacting to this new benchmark of frontier math where LLMs only solve 2%. It was introduced because LLMs are increasingly crushing existing math benchmarks. The interesting issue is that even though by many accounts (/evals), LLMs are inching…

1/10 Today we're launching FrontierMath, a benchmark for evaluating advanced mathematical reasoning in AI. We collaborated with 60+ leading mathematicians to create hundreds of original, exceptionally challenging math problems, of which current AI systems solve less than 2%.

Tweet Image 1


Geo.vanny Reposted

Since our first version of Hermes was released over a year ago, many people have asked for a place to experience it. Today we're happy to announce Nous Chat, a new user interface to experience Hermes 3 70B and beyond. hermes.nousresearch.com We have reasoning enhancements, new…

Tweet Image 1

Geo.vanny Reposted

The 3-3-3 Method

Tweet Image 1

Geo.vanny Reposted

Best practices for data visualisation:

Tweet Image 1

Geo.vanny Reposted

Tips for Writing a Research Paper using LaTeX github.com/guanyingc/late…

Tweet Image 1

Geo.vanny Reposted

That’s wild

NEWS: NVIDIA and xAI have unveiled Colossus, the world’s largest AI supercomputer, built in just 122 days and powered by NVIDIA’s groundbreaking Spectrum-X Ethernet networking. “Colossus is the most powerful training system in the world,” said Elon Musk on X. “Nice work by the…

Tweet Image 1


Geo.vanny Reposted

Hippies: que es más importante el arte o la vida 😭😭😭 El dueño del museo: no me rompa los huevos pibe


Geo.vanny Reposted

Open-source AI is the key to unlocking the widespread potential of artificial intelligence, democratizing the technology, and putting it in the hands of the many rather than the privileged few. On Safety: nousresearch.com/from-black-box…


Geo.vanny Reposted

Adam Optimizer's Bias-Correction ~ Math vs Code 🔢💻 ~ Continuing from the previous visualizations on the first and second moment estimates, I made this visualization to show you how to implement the bias-correction steps in the Adam Optimizer using PyTorch within 2 LoC. Bias…


Geo.vanny Reposted

The Fourier Transform, explained in one sentence by Stuart Riffle. [bityl.co/NGqj]

Tweet Image 1

Geo.vanny Reposted
Tweet Image 1

Geo.vanny Reposted

Google Deepmind trained a grandmaster-level transformer chess player that achieves 2895 ELO, even on chess puzzles it has never seen before, with zero planning, by only predicting the next best move, if a guy told you "llms don't work on unseen data", just walk away

Tweet Image 1

United States Trends
Loading...

Something went wrong.


Something went wrong.