paul @binaryfoundry

Developing a real time ray tracing framework. https://t.co/aYJGg02kyR binaryfoundry.github.io Joined March 2014

Tweets

573
Followers

129
Following

423
Likes

3K

Leon @LeonWohlgemuth

4 weeks ago

Beautiful open-source WebGL fluid simulation by @PavelDoGreat. An interactive demo that computes fluid dynamics and visualizes evolving flow fields. Try it here (code in the comments): paveldogreat.github .io/WebGL-Fluid-Simulation

15 53 508 18K 231

Download Video

Xor @XorDev

2 years ago

"Origami" #つぶやきGLSL o++;vec4 h;vec2 u;for(float A,l,a,i=.6;i>.1;i-=.1)a-=sin(a-=sin(a=(t+i)*4.)),u=(FC.xy*2.-r)/r.y,l=max(length(u-=rotate2D(a/=4.)*clamp(u*rotate2D(a),-i,+i)),.1),A=min((l-.1)*r.y*.2,1.),o=mix(h=sin(i/.1+a+vec4(1,3,5,0))*.2+.7,o,A)*mix(h/h,h+.5*A*u.y/l,.1/l);

12 100 935 58K 212

Download Video

Amit Patel @redblobgames

4 weeks ago

Love this blog post from @imadr_

dek @shatterspine

4 weeks ago

Love this blog post from @imadr_

1 25 219 19K 148

Download Video

2 14 129 9K 57

Xor @XorDev

a month ago

On branching in shaders: If-statements are generally better than math operations. The concerns about branching are mostly misplaced based on very old hardware. The main focus should be on avoiding divergence: pixels near each other should do similar things. Branching can become a…

20 13 325 21K 106

Leon @LeonWohlgemuth

a month ago

Awesome visualization shows hundreds of AI completions of the prompt "Intelligence is..." in 3D space, showing how responses branch over time. Try it here (by Santiago Ortiz): moebio .com/mind/

17 107 884 62K 522

Download Video

eigenron @eigenron

2 months ago

Llama-2-7B with KV-cache, RoPE, and GQA in 240 lines of code (without subtracting comments).

4 17 173 11K 114

Download Image

maharshi @mrsiipa

2 months ago

this was the simplest explanation of tensor and expert parallelism i have come across so far

5 105 2K 74K 2K

Download Image

ℏεsam @Hesamation

2 months ago

learn how to build an LLM from scratch, honestly. @rasbt's repo is really a gem. it has notebooks with diagrams and explanations that will teach you 100% of: > attention mechanism > implementing a GPT model > pretraining and fine-tuning my top recommendation for studying LLMs.

31 510 3K 219K 5K

Download Image

eigenron @eigenron

2 months ago

i saw this guy’s video on simulating gravity in C++ and that pushed me into learning OpenGL (failure lol). now he’s back with another one simulating black holes. he incorporated ray tracing, Einstein’s field equation, Schwarzschild solution to the field equation, space-time,…

70 611 9K 276K 4K

Download Image

Leon @LeonWohlgemuth

2 months ago

Watch how an LLM processes a word, step by step. Great for visual learners. Clean, interactive, and super informative. Source: bbycroft .net/llm

10 32 195 8K 177

Download Video

karan @karancl

2 months ago

transformers for everyone, if you can read english let this be your introduction to deep learning, you will not regret this!!!

14 236 3K 154K 3K

Download Image

Grant Sanderson @3blue1brown

2 months ago

New video on the details of diffusion models: youtu.be/iv-5mZ_9CPY Produced by @welchlabs, this is the first in a small series of 3b1b this summer. I enjoyed providing editorial feedback throughout the last several months, and couldn't be happier with the result.

40 410 3K 368K 1K

Pramod Goyal @goyal__pramod

2 months ago

A short mathematical blog on Softmax It's more interesting than it looks

2 65 535 22K 417

Download Image

Xor @XorDev

2 months ago

"Qubit" vec3 p,v;for(float i,z,d;i++<5e1;o.rgb+=v/d/z)p=z*normalize(FC.rgb*2.-r.xyy),p.z+=6.,v=max(v=abs(p=dot(v=normalize(cos(t*.5+vec3(0,2,4))),p)*v+cross(v,p)),max(v.y,v.z)),z+=d=.2*abs(v.x-2.)+.1*length(min(v=abs(p)+sin(p/v.x*6.+t+t),v.yzx)-1.);o=tanh(o/2e2);

12 87 624 13K 131

Download Video

Pramod Goyal @goyal__pramod

3 months ago

Understanding VAE by the creators of VAE

5 89 891 51K 904

Download Image

Ben Sims @BenSimsTech

3 months ago

How to PBR: n1(air ior): ~1 n2(dielectric ior): 1.5 f0(specCol) = ((n1-n2)/(n1+n2))^2 == 0.04 f0 = lerp(f0, albedo, metallic) fresnel = lerp(f0, 1, pow(1-LdotH, 5)) spec = GgxDv(roughness)*f0 diffuse = lerp(albedo, 0, metallic) return (specular + diffuse)/Pi * NdotL * lightColor

3 7 92 6K 70

Pramod Goyal @goyal__pramod

3 months ago

Some of these were recent blogs on attention and transformers I read that made the idea so much more intuitive and simple to me

Pramod Goyal @goyal__pramod

3 months ago

Some of these were recent blogs on attention and transformers I read that made the idea so much more intuitive and simple to me https://t.co/wjmLk8EaCN