RuiningLi

Introducing Particulate: a feed-forward model for 3D object articulation 💻✂️👓🧳 Particulate gives you a fully articulated 3D object, including part segmentation, kinematic structure & motion constraints, in a single forward pass in ~10secs. 🏅SOTA performance! 💡GenAI compatible: Turns AI-generated 3D meshes into fully articulated models! Project page: ruiningli.com/particulate Code: github.com/RuiningLi/part…

@acatovicx Not as good as 5.5 and Gemini yet :((

@RayLi234 This is awesome - instead of creating bespoke models you use latest LLMs and surround it with a 3d asset generating harness. It also looks fairly cheap, sub $1.50 per asset generation. Would be interesting to know how well Qwen and open models work?

🚀 Introducing Articraft, a coding agent for articulated 3D asset creation. Articraft writes code, executes it, receives validation feedback, and refines the result into simulation-ready 3D assets with parts, joints, and motion. We’re also releasing Articraft-10K: 10,000+ articulated objects across 250 categories, unlocking large-scale interactive scenes for robotics simulation and physical AI. 🔗 Project page: articraft3d.github.io 💻 Code: github.com/mattzh72/artic…

@NicholasEPfaff Thanks! We don’t have a lot of money so scaling to 10k assets requires the agent to be cheap. @Mattzh1314 made lots of tradeoffs between fidelity/realism and cost. A big fan of your work Scenesmith, we should integrate these pipelines together for more scalable real to sim!

@RayLi234 very cool! I love that you get these results without any visual feedback and that the system is relatively fast/cheap

This is an amazing project led by @Mattzh1314. I thought LLMs could already code 3D, but I did not expect such realistic, complex, and fine-grained articulation generation. It completely changed my view of 3D agents — this could be a major step toward real industrial agents.

When we started this project, the goal was to augment training data for Particulate (ruiningli.com/particulate). Then @Mattzh1314 kept adding magic to the agent, and at some point we were debating: do we still need neural generators for 3d design, or should we go all in on agents?

Check out Ariticraft 🦾 - a highly efficient agentic system that generates articulated 3D assets fully automatically at scale! 🚀 articraft3d.github.io

@Mattzh1314 @XiaoyangLyu22 🔊We are welcoming data contribution from the community! You can share the assets you generated with everyone by simply creating a pull request (check #data-contribution-workflow" target="_blank" rel="nofollow noopener">github.com/mattzh72/artic…). We will actively maintain the dataset and can't wait to see what people build with the assets!

Check out our project page (articraft3d.github.io) for more examples! The code is open-sourced with Apache-2.0 at github.com/mattzh72/artic…. The project is led by @Mattzh1314 (I learned so much on coding agents from him), with incredible contributions from @XiaoyangLyu22, Matt Song, @ZheningHuang and an amazing advisor team @ChuanxiaZ, Christian, Andrea and @elliottszwu.

@baaadas @LumaLabsAI Looks amazing! How could we try?

Excited to introduce Uni-1, our new *unified* multimodal model that does both understanding and generation: lumalabs.ai/uni-1 TLDR: I think Uni-1 @LumaLabsAI is > GPT Image 1.5 in many cases, and toe-to-toe with Nano Banana Pro/2. (showcase below)

@wzhao_nlp What are some great resources for learning ML infra as an ML researcher?

🌶️ Some (perhaps) spicy thoughts. It’s been a while since my last tweet, but I wanted to write about how disorienting it has been from academia to an LLM lab 😅 The kind of research I was trained to do during my PhD almost doesn’t exist here. The obsession with mathematical elegance and novelty is mostly gone. Everything is about scaling data and compute. For a while, that really got to me. At my lowest point, I felt like I’d lost interest in building LLMs altogether. I didn’t feel intellectually challenged anymore. What made this even stranger was that, at a technical level, things worked. If there was a capability I wanted to teach a model, scaling the right data and compute always got me there, no exception (so far). But recently, I found a way to reconcile with myself.. I realized the real competition isn’t in the ML recipe anymore. Most teams do roughly the same thing. What actually matters is how fast you can iterate, test ideas, and recover from mistakes. And that speed is mostly backed by infrastructure 🏗️ Faster loops, fewer bugs, better tooling. Seeing this made me excited again! Infra is its own deep, hard, and intellectually fun problem space. In 2026, I want to become an ML researcher who’s really good at infra. And I'll come back to ML problems with that edge, and will be excited to share what I find 😌

@GeminiApp Gemini has been great! But a real game changer that keeps me coming back to GPT is its ability to execute code directly. Any plans to bring something like that to Gemini?

Gemini 3 Flash is here ⚡️ Get the free, fast, unlimited version of Gemini with our biggest upgrade yet. See what’s new and how we’re using Gemini 3 Flash to tackle everyday tasks and get answers fast. (thread)

Work co-led by Yuxin from @Cambridge_Uni, with the amazing team: @ChuanxiaZ, Christian, Joan, @elliottszwu, and Andrea from @Oxford_VGG. More 3D results on the project page: ruiningli.com/particulate.

From a static 3D mesh (either crafted or generated), the model samples a surface point cloud and directly infers part segmentation, kinematic structure, and motion constraints—fully parameterizing an articulated object ready for URDF-based simulation.

Introducing Particulate: a feed-forward model for 3D object articulation 💻✂️👓🧳 Particulate gives you a fully articulated 3D object, including part segmentation, kinematic structure & motion constraints, in a single forward pass in ~10secs. 🏅SOTA performance! 💡GenAI compatible: Turns AI-generated 3D meshes into fully articulated models! Project page: ruiningli.com/particulate Code: github.com/RuiningLi/part…

Particulate Feed-Forward 3D Object Articulation