chrisdeister

Not a fan of Woodrow Wilson.

[Aakash Gupta @aakashgupta]

Yann LeCun closed $1.03B for AMI Labs on March 10. Three days later, this paper dropped from his NYU collaborators.

15M parameters. Single GPU. A few hours of training.

LeWorldModel is the first JEPA that trains end-to-end from raw pixels. Two loss terms: predict the next embedding, keep the latent space Gaussian. Previous JEPAs needed exponential moving averages or pretrained encoders to avoid representation collapse. LeWM doesn't.

Six hyperparameters down to one.

The numbers are the story. Foundation-model-based world models require hundreds of millions of parameters and serious compute to plan a control task. LeWM plans up to 48x faster while staying competitive on 2D and 3D benchmarks. The whole thing fits on a laptop GPU.

Look at the trajectory. Yann announced his Meta departure in November 2025 after 12 years and called founding FAIR his "proudest non-technical accomplishment." On March 10, 2026, AMI Labs closed the largest seed round in European history at a $3.5B pre-money valuation. Bezos, Nvidia, Samsung, and Toyota all wrote checks.

Three days later: a paper showing that JEPA-from-pixels is no longer fragile and no longer compute-heavy. The engineering scaffolding that made it look like an academic curiosity is gone.

The authors sit at Mila, NYU, Samsung SAIL, and Brown. None at Meta.

Yann's bet was that the path to machine intelligence runs through world models, not language models. He left a public company to build it. Each JEPA paper from his network resets the assumed cost structure for that bet. This one makes world modeling laptop-cheap.

Meta still has the GPUs. The architecture left.

[Mac Baconai @Macbaconai]

have a great day

[Oscar Arias @OACerebro]

Las figuras deben ser lo más simples posibles.

Figura 1:
a) Un diagrama demasiado complicado de una inversión en dos genes.

b) Una versión simplificada, combinando los dos primeros pasos y usando menos flechas.

[.stuff @vintagestuff4]

Manaus, Brazil

[Frank Nielsen @FrnkNlsn]

👀👀👀

"Geometric bifurcation theory: Fisher information geometry applied to dynamical and complex systems"

[tetsuo @tetsuoai]

λ-2D: Drawing as Programming Language, Featuring Ideas from Lambda Calculus.

MIT Media Lab

[Dilum Sanjaya @DilumSanjaya]

Fun interactive science app ideas | Part 3

Played around with generating 3D biological structures and made an app to explore them interactively

UI Design
GPT Images 2

Code
Gemini 3.1 Pro

More demos ↓

[Antonio Lupetti @antoniolupetti]

"Algebrica" is a free and open mathematical knowledge base. All entries are progressively being released in Markdown format on GitHub for anyone who wants to study mathematics freely and openly.

Alongside the texts, the individual SVG illustrations are also made freely available. They are minimal, mathematically accurate, and designed to be easily reusable in notes, lecture material, or educational resources. Since they are vector-based and code-driven, they can also be modified or improved simply by editing the source.

Another step toward making the knowledge base more open, transparent, and genuinely useful over time.

[Kirk Borne @KirkDBorne]

Problem Solving Through Recreational Mathematics (Math Games & Puzzles): https://t.co/YhLWnMy6aF

[ParametricArchitecture @parametricarch]

The Quiet Power of White 🌫️

[Jürgen Schmidhuber @SchmidhuberAI]

World models rock!

JS. The Neural World Model Boom. Technical Note IDSIA-2-26, 2/4/2026, updated 4/29/2026 https://t.co/bRz8PFNt0P

[ParametricArchitecture @parametricarch]

Floating Geometry on Jinji Lake 🌊

[Shunzhi Wang @wang_shunzhi]

1/ Excited to share our new Review in @Nature:
“The past, present and future of de novo protein design.”
Here, we mainly focused on structure-guided protein design. The field is entering a new phase: now that we can design new proteins, what should we build next?

[Ivan Marti Vidal @imartividal]

Esta figura muestra datos de un gravímetro cuántico (que mide la fuerza de la gravedad con altísima precisión) instalado en el Zugspritze (la montaña más alta de Alemania). Aquí vemos cómo cambia la gravedad con el tiempo, debido a efectos muy interesantes. ¡Hilo!

[Antonio Lupetti @antoniolupetti]

In AI, causal learning without backpropagation means learning causal direction through local interactions instead of gradient-based training.

This paper moves away from gradient-based training and shows how causal direction can emerge from neural assemblies through local plasticity and sparse co-activation.

Directionality is not inferred after training, but reflected in asymmetric synaptic structure shaped by local interactions. In controlled settings, the model recovers the underlying causal graph.

It feels less like optimization and more like structure emerging from local interactions.

https://t.co/7YdyiFpPQ4

[Mathematica @mathemetica]

This periodic table is sized by how abundant each element is on Earth.

Look how tiny Helium (He) is... yet we fill party balloons with it like it’s unlimited.
Helium is irreplaceable for MRIs, superconductors, aerospace, and cryogenics… and once it escapes into space, it’s gone forever. We should treat it as the precious resource it is.