Lahia.o

Typing on the keyboard ... it's work 🏴

[Brett Adcock @adcock_brett]

Watch a team of humanoid robots running a full 8-hr shift at human performance levels. This is fully autonomous running Helix-02 https://t.co/IdZR0T1F5I

[Lahia.o @df_trainX]

The word ‘mean’ has a funny double life in English.

In math, the ‘mean’ is the average, the central value of numbers.

In language, meaning is the central idea of words.

Different origins but the same intuition, finding the center of something.

[Lahia.o @df_trainX]

@Tahabsn Is this a visualization of some NN weights?

[Lahia.o @df_trainX]

@laytoun I think It a human reflex( learn while training), when we are in an open community, we tend to create internal protocol/language/… to hide ourselves from the outside world.

[Extropic @extropic]

Thermodynamic Intelligence: First Contact

10.29.25

[Lahia.o @df_trainX]

A prompt is your sail on the ocean of LLMs parameters. With the right one, you reach your destination, the desired output. With the wrong one, you drift endlessly through the wrong tokens.

#AI

[Lahia.o @df_trainX]

@jasonzhou1993 @cursor_ai @cline Same thing with openai api 🥲

[Lahia.o @df_trainX]

Thanks to @gabriel1 for the idea on one of his tweet 😁

[Lahia.o @df_trainX]

🌀 Maze-Diffusion — I trained a diffusion model to solve mazes!
No A*, no Dijkstra… just generative pathfinding with ✨
🧩 Conditioning: Maze layout + start/end
🤖 Model: U-Net + DDPM
🛠 Built with PT, HF & ❤️

🔗 Try it live: https://t.co/tA2bNWWF4u
#AI #DiffusionGenerativeAI

[Lahia.o @df_trainX]

It's all about encoding and decoding, who is doing it right, win.

[Lahia.o @df_trainX]

@smakosh @bryan_johnson Nice one 🤣 level s with 100 score sleep

[Lahia.o @df_trainX]

Our languages use every word as a class, a container of meaning and feeling. Each word captures a subset of the complexity of nature and compresses it to make this complexity easier for us to be interpreted.

[Lahia.o @df_trainX]

nature has no labels, it's a feature we create for us.

[Lahia.o @df_trainX]

Human written and spoken languages serve as systems of classification, allowing our brains to organize and interpret the environment, helping us make sense of the world around us.

[Lahia.o @df_trainX]

We need an AI/model that can create and update its own weights
#AI

[Lahia.o @df_trainX]

Ouuff 🔥🔥

[colinhacks/zod @colinhacks]

sometimes I write code that weaponizes the insanity of JavaScript just to feel alive

[∩ @zachpogrob]

Bukowski

[Lahia.o @df_trainX]

@kyutai_labs The response time is VERY FAST 🔥, is that why ' Moshi thinks and speaks at the same time ' ??

[François Chollet @fchollet]

I'm partnering with @mikeknoop to launch ARC Prize: a $1,000,000 competition to create an AI that can adapt to novelty and solve simple reasoning problems.

Let's get back on track towards AGI.

Website: https://t.co/wNsM3IQgEI

ARC Prize on @kaggle: https://t.co/Lhsh1RiWKq