🔥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

Meet #Hillbot Alpha, the first fully autonomous mobile manipulation robot trained using sim-to-real technology. Designed in Hillbot’s San Diego headquarters, Hillbot Alpha represents the potential of data synthesis via simulation in robotics. #AGI #EmbodiedAI #AI #Sim2Real

🚀🤖 Top 10 Robot Learning Papers of 2024 is out!🔥 With 2 rounds of nominations & voting, plus 330+ individual votes, these standout papers shine across diverse categories: 1️⃣ π0: Vision-Language-Action Flow Model for General Robot Control 2️⃣ Closed-Loop Open-Vocabulary Mobile

Accepted to ICLR 2025! :D

Excited to share that I’ll be joining UC San Diego for my PhD, advised by Professor Hao Su (Hao Su)! Many thanks to everyone who helped me along my research journey so far — I’m looking forward to continuing research in robot learning, manipulation, and simulation!

ManiSkill3 accepted as an oral paper at the ICLR 2026 robot learning workshop! See you all in Singapore🇸🇬, we have some cool demos demonstrating all the power of parallel rendering, heterogeneous sim and more! We will also update the arxiv with a v2 with some of these updates

Can SAC achieve fast wall-time with GPU-parallelized sim? Can SAC beat PPO, the go-to RL algo for massively parallel envs? Yes! Using simple implementation tricks and tuning, SAC can match and beat tuned PPO+CudaGraphs baselines in ManiSkill wall-time: arthshukla.substack.com/p/speeding-up-…

🚀Excited to release PartField—a feedforward model that learns part-based feature fields for 3D shapes! It enables lightning-fast⚡️, robust, open-world hierarchical 3D part seg and unlocks cross-shape applications like co-seg and correspondence! 🔗shorturl.at/HnUmc 1/n

I'll be at #ICLR2025 next week to present ManiSkill-HAB. I work on the intersection of robot learning, ML/RL, and sim. If you'd like to meet up and chat, please let me know!

🤔 How to fine-tune an Imitation Learning policy (e.g., Diffusion Policy, ACT) with RL? As an RL practitioner, I’ve been struggling with this problem for a while. Here’s why it’s tough: 1️⃣ Special designs (usually for multimodal action distributions) in modern IL models make

Presenting ManiSkill-HAB today, a holistic benchmark for low-level manipulation in home rearrangement tasks! Stop by poster 28 at 3:00PM GMT+8 today (poster session 2) to chat about RL, manipulation, sim, and open source robotics!

I’ll be at the robot learning workshop today and giving an oral talk at 9:15 AM on ManiSkill3 in rooms Garnet 216/217 at ICLR. Come see the crazy things you can do with fast sim+rendering like fast visual RL and zero-shot RGB sim2real! ManiSkill was also accepted at RSS!

Had a great time at #ICLR2025 last week! Met lots of new people, and Singapore was beautiful.

Really cool mobile manip teleop in ManiSkill! So many cool SO-100 projects, feeling like I should get one for myself…

🧠 Can a single robot policy control many, even unseen, robot bodies? We scaled training to 1000+ embodiments and found: More training bodies → better generalization to unseen ones. We call it: Embodiment Scaling Laws. A new axis for scaling. 🔗 embodiment-scaling-laws.github.io 🧵👇