🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Our CoRL 2024 paper shows Reinforcement Learning can allow robots to learn skills via real-world practice, without any demonstrations or simulation engineering. Rewards are provided using language/vision models, and mobility of robots enables autonomous exploration. 1/N

Excited to release RT-Affordance! We propose conditioning policies on visual affordance plans as an intermediate representation that allows us to learn new tasks without collecting any new robot trajectories. Website and paper: snasiriany.me/rt-affordance Here’s a short 🧵

Astribot 🤝 Physical Intelligence (π)

So impressed with the speed & ease of use of MJX + high-quality real-world transfer for all these different tasks! Being able to change something in sim and see the effects almost immediately in the real world is so powerful. Congrats!

So exciting to see such great progress on MJX! I've been waiting for this for so long! Here's my personal experience: we started a chat after Kevin Zakka began training, and before the chat even ended, Kevin already had a training result that could be deployed on the robot!🤯

We've just added the Booster Robotics T1 Humanoid to MuJoCo Playground, with code to train a joystick controller as well as a deployable ONNX file. A super fun weekend with Carlo Sferrazza and Qiayuan Liao :) github.com/google-deepmin…

Introducing Hi Robot – Hierarchical Interactive Robot Our first step at Physical Intelligence towards teaching robots to listen and think harder. A 🧵 on how we make robots more steerable 👇

My knee-jerk reaction was that sim-2-real wouldn't work watching the learned policy in sim (on left), but actually it does!

Physical Intelligence we trained π-0.5, a model that can do tasks in homes never seen in training! Here is one video of π-0.5 completing a long-horizon task in an unseen kitchen while I provide natural language instructions. A short thread 🧵

How to build vision-language-action models that train fast, run fast & generalize? In our new paper, we formalize & analyze the approach of our π-0.5 model & further improve it with a single stage recipe. Blog: pi.website/research/knowl… Paper: pi.website/download/pi05_…