Neural CAs are amazing, but they've never scaled past low resolution. We propose a simple solution that allows an ~8x resolution boost with minimal extra parameters. The core idea: Treat cells as local neural fields instead of pixels. Try the demo: cells2pixels.github.io/ 🧵

the hardware lottery is also the hardware curse. with our team, we've been pushing the boundaries of what fully local neural cellular automata models achieve on visual reasoning. we believe this opens the door to a swathe of efficient, robust, and scalable hardware designs.

Can complex multi-step reasoning emerge purely from cells that only talk to immediate neighbors? Happy to share our paper “Reasoning with Neural Cellular Automata (NCA)”, from our team at Google, Paradigms of Intelligence 🧵👇

Thanks to AI I make 10x more failed experiments now!

a=(m,d=mag(k=9*cos(i*5)*sin(i),e=cos(i*3)*cos(i*2)*9)**3/1999+1.5-sin(t/2+m)**3/3)=>point(99*sin(c=d/16-t/48+m)+k*(p=d**sin(d*d-t+m))+200,99*sin(c*4)+e*p+200) t=0,draw=$=>{t||createCanvas(w=400,w);background(9).stroke(w,96);for(t+=PI/20,i=1e4;i--;)a(i%16*13)}//#つぶやきProcessing

You've heard of Langton's ant. Check out its trainable version. Since it borrows from the architecture of NCAs should we call it Mordvintsev's ant? #ALIFE2026 #ERAWorkshop

New ALife late-breaking abstract with @drmichaellevin! We find that lifelike beings with species-specific reactions to perturbations emerge from Gifbreeder despite the absence of explicit agents, environments, or local interactions. I welcome any feedback on these early ideas.

Starting to explore forms possible in @zzznah's MorphoHDL, using CLIP w/ evolutionary search based on a prompt. (As you can see, Fable was excited to name them and do the visuals 😂)

Always wanted to try combining my paper with NCA, and turns out this approach is a perfect starting point It decouples grid with rules from rendering by using SIREN as a decoder, so I just replaced it with SAD Same rules, different way to render them in higher resolution

Can't believe NVIDIA showed this world model demo which is full of hallucinations in their keynote... #SIGGRAPH2026

Model organism with NCA. Based on @zzznah original work and @Esychology “from cell to pixels”

Who is it?

Introducing MorphoHDL, a minimal language prototype for growing boolean circuits! paradigms-of-intelligence.gi… Early this year I wanted to design a size-agnostic graph rewrite rule system that could build functional boolean circuits. First I thought about "chemistry"-like reactive systems, but rules were too complicated with many different node types carrying multiple indices... Then I realized, that cell division is much more natural way of building complex structures, and once I started to treat graph edges as buses instead of single wires, everything clicked. No new formalism — just taking good old recursion and seeing how far I can push it. Ripple-carry adder, Brent–Kung adder, multipliers ... and some Haeckel-esque creatures along the way! Hope you enjoy the report of my journey and the demos.

My whole life experience screams that the correlation between effort and lines of code must be negative

•◽️• @jn3008 + @etiennejcb

41 years ago, Carl Sagan gave a clear and elegant explanation of the science behind global warming.