📢Delighted to share Recurrent Interface Network (RIN), a new model/architecture that improves performance of state-of-the-art UNet-based image and video diffusion models (on pixels) while achieving 10X efficiency gain. arxiv.org/abs/2212.11972 Joint work w Allan Jabri, David Fleet

🎯 Modeling how people interact with objects in #3D is important for both 3D #perception from 2D images, and for 3D #avatar synthesis for AR/VR 👉 I gave a talk on this @ #BMVA 'Symposium on Digital Humans' 👉 Video: youtu.be/z-Q7DcSL_vE 👉 Slides: dropbox.com/scl/fi/y7wgft8…

The bane of my existence is the so-called "Modern" Search Engines that are incapable of finding a fucking substring in a string.

Тsфdiиg If only substring matching had been a well-studied problem in computer science, someone might have come up with a good algorithm for this problem that github could implement!

Power-Linear Polar Directional Fields Jiabao Brad Wang, Amir Vaxman arxiv.org/abs/2506.00222 Abstract: We introduce a novel method for directional-field design on meshes, enabling users to specify singularities at any location on a mesh. Our method uses a piecewise power-linear

If you’re interested in learning about Continuous Thought Machines (sakana.ai/ctm/), we made interactive notebook tutorials so you can hack around with CTMs ImageNet: github.com/SakanaAI/conti… MNIST Tutorial: github.com/SakanaAI/conti… Let me know if you have any feedback!

So Flow Matching is *just* xt = mix(x0, x1, t) loss = mse((x1 - x0) - nn(xt, t)) Nice, here it is in a fragment shader :) shadertoy.com/view/tfdXRM

Demystifying the Nvidia Ampere Architecture through Microbenchmarking and Instruction-level Analysis arxiv.org/pdf/2208.11174

GPU Programming Primitives for Computer Graphics gpu-primitives-course.github.io/sa-course-note…

Globally Consistent RGB-D SLAM with 2D Gaussian Splatting Xingguang Zhong, Yue Pan, Liren Jin, Marija Popović, Jens Behley, Cyrill Stachniss tl;dr: 2DGS->map; rendering-based frame-to-map camera pose estimation; MASt3R->loop closure arxiv.org/abs/2506.00970

E3D-Bench: A Benchmark for End-to-End 3D Geometric Foundation Models Wenyan Cong, Yiqing Liang, Yancheng Zhang, Ziyi Yang, Yan Wang, Boris Ivanovic, Marco Pavone, Chen Chen, Zhangyang Wang, Zhiwen(Aaron) Fan tl;dr: in title arxiv.org/abs/2506.01933

TLDR, you can speed up training by initializing some layers with dinov2 layers instead of random weights

A deterministic multimodal counterpart of particle filters. A new class of state space model in moment representation for optimal estimation with arbitrary noise density. Accepted for Ro

Today, we are introducing SmolVLA: a 450M open-source vision-language action model. Best-in-class performance and inference speed! And the best part? We trained it using all the open-source LeRobot datasets in the Hugging Face hub! But how? 🫳🏀

We have a new SSM theory paper, just accepted to COLT, revisiting recall properties of linear RNNs. It's surprising how much one can delve into, and how beautiful it can become. With (and only thanks to) the amazing Alexandre and Francis Bach arxiv.org/pdf/2502.09287

UniWorld is out on Hugging Face High-Resolution Semantic Encoders for Unified Visual Understanding and Generation

Most video models 🤯forget the past 🐌slow down over time 🔁rely on bidirectional (not causal) attention Our state-space video world models (SSM) 🧠remember across hundreds of frames ⚡️generate at constant speed ⏩is fully causal, enabling real-time rollout 1/3

my NGJ2025 talk about Quaternions is up! 💫 youtube.com/watch?v=PMvIWw…

AI that can improve itself: A deep dive into self-improving AI and the Darwin-Gödel Machine. richardcsuwandi.github.io/blog/2025/dgm/ Excellent blog post by Richard C. Suwandi reviewing the Darwin Gödel Machine (DGM) and future implications.