Justin Yu ✈️ CoRL

This account is verified. Learn more

@uynitsuj

PhDing @UCBerkeley @berkeley_ai advised by Prof. @Ken_Goldberg. Robotics, AI | Intern @xdofai

Berkeley, CA uynitsuj.github.io Joined June 2023

237 Following 471 Followers 132 Posts

Latest posts 5

Justin Yu ✈️ CoRL @uynitsuj Oct 27

Did not expect to see our Real2Render2Real coffee maker making rounds on twitter today.

Glad this video I captured over a year ago now is still contributing to scaling the real2sim2real agenda for manipulation!

Rundong Luo Verified account @LuoRundong0122 Oct 27

1/7 Excited to share SceneRig! 🦾

One image → an editable, simulation-ready 3D scene.

We combine AI agents with geometry and physics to reconstruct scenes for robot policy evaluation and replay of real-world manipulations.

Here’s how it works 🧵👇

View quoted post on X

Justin Yu ✈️ CoRL @uynitsuj Sep 12

WARP-RM (Warp Augmented Relative Progress Reward Model) accepted to CoRL 2026 🎉 Code is public and I invite you to reproduce our open-data simulation results end-to-end 💻

I also invite folks with more API credits to zero-shot or ICL this task w/ Astra or Fable 😎 Very curious if current models can succeed, and if so, at what time & token cost

Justin Yu ✈️ CoRL Verified account @uynitsuj Jun 30

Introducing WARP-RM: A Warp-Augmented Relative Progress Reward Model for Data Curation 🧵

We gave our t-shirt folding robot more demonstrations and it got worse. Every extra demo ended in a successfully folded shirt. The data wasn't bad. It was noisy. The policy couldn't tell productive motion from dead time, and it imitated both equally. So which moments of a demo are actually worth copying?

🌐 Project Website: uynitsuj.github.io/warp-rm
📄 Paper: arxiv.org/abs/2606.28320
💻 Code: github.com/uynitsuj/WARP-…
📨 XDOF blog post: xdof.ai/blog/warp-rm

View quoted post on X

Justin Yu ✈️ CoRL @uynitsuj Jul 15

Real2Render2Real-pilled mobile manipulation :D Kudos to @Lingxiao234 and @imhuanyuli for the brilliant idea and execution.

lingxiao guo ✈️CoRL2026 Verified account @Lingxiao234 Jul 14

Reality of mobile manipulation: 1,000 demos buy you 1 task, in 1 scene.

Introducing 𝐖𝐀𝐍𝐃𝐀: a synthetic data engine that turns ONE demo into thousands of trajectories across unseen scenes — and the policy works in the open world.

One demo in. An open world out.

🧵

View quoted post on X

Justin Yu ✈️ CoRL @uynitsuj Jun 30

Excited to be attending ICML this year ✈️🇰🇷 and presenting CaP-X on agentic robotics (co-led with @letian_fu and @HaoruXue, among others who are not on X!)

DMs are open. Let’s grab a coffee and talk shop on harnesses for embodied AI, reward modeling, or the latest in hardware!

Max Fu Verified account @letian_fu Apr 1

Robotics: coding agents’ next frontier.

So how good are they?

We introduce CaP-X: an open-source framework and benchmark for coding agents, where they write code for robot perception and control, execute it on sim and real robots, observe the outcomes, and iteratively improve code reliability.

From @NVIDIA @Berkeley_AI @CMU_Robotics @StanfordAILab

capgym.github.io

🧵

View quoted post on X

Justin Yu ✈️ CoRL @uynitsuj Jun 30

Big thanks @QianzhongChen ofc standing on the shoulders of giants such as your work on Stage Aware Reward Modeling arxiv.org/abs/2509.25358 😎😎

arxiv.org SARM: Stage-Aware Reward Modeling for Long Horizon Robot Manipulation Large-scale robot learning has made progress on complex manipulation tasks, yet long horizon, contact rich problems, especially those involving deformable objects, remain challenging due to...

Qianzhong Chen ✈️ CoRL2026 Verified account @QianzhongChen Jun 30

Big big step from value model to direct reward/adv. model. People in this area know how hard it is, congrats @uynitsuj and the team!
View quoted post on X