Reasoning Models Thinking Slow and Fast at Test Time Another super cool work on improving reasoning efficiency in LLMs. They show that slow-then-fast reasoning outperforms other strategies. Here are my notes:

Thank Akshay for sharing! I will post an introduction to our new work, AlphaOne, soon! Stay tuned!

Very impressive results! I want my G1 to serve me a beer as well🍻

Can you imagine playing various games through an AI model? Like BlackMyth: Wukong.🤩 Sharing our latest work: DeepVerse, an autoregressive paradigm-based world model🌏DeepVerse can fantasize the entire world behind images and enable free exploration through interaction🎮.

This is 🤯 Figure 02 autonomously sorting and scanning packages, including deformable ones. The speed and dexterity are amazing.

Motion tracking is a hard problem, especially when you want to track a lot of motions with only a single policy. Good to know that MoE distilled student works so well, congrats Zixuan Chen on such exciting results!

#RSS2025 Excited to be presenting our HumanUP tomorrow at the Humanoids Session (Sunday, June 22, 2025) 📺 Spotlight talk: 4:30pm–5:30pm, Bovard Auditorium 📜Poster: 6:30pm-8:00pm, #3, Associates Park

AGIBOT X2-N (Nazhe) new video Shows the ability to carry goods blindly on stairs and slopes📦 The robot autonomously switches between bipedal and wheeled modes while maintaining balance and stability throughout the process — a feature that will be highly valuable in dim or

🤖 What if a humanoid robot could make a hamburger from raw ingredients—all the way to your plate? 🔥 Excited to announce ViTacFormer: our new pipeline for next-level dexterous manipulation with active vision + high-resolution touch. 🎯 For the first time ever, we demonstrate

DreamVLA A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge

🔥 Thrilled to release our new multimodal RL work: Open Vision Reasoner! A powerful 7B model with SOTA performance on language & vision reasoning benchmarks, trained with nearly 1K steps of multimodal RL. Our journey begins with a central question: Can the cognitive behaviors

Everyone knows action chunking is great for imitation learning. It turns out that we can extend its success to RL to better leverage prior data for improved exploration and online sample efficiency! colinqiyangli.github.io/qc/ The recipe to achieve this is incredibly simple. 🧵 1/N

Really cool invited talk by Shuran Song - “Making Video Model Useful for Robots”

Getting up is nothing fancy nowadays but it still feels great to have it.

How do we unlock the full dexterity of robot hands with data, even beyond what teleoperation can achieve? DEXOP captures natural human manipulation with full-hand tactile & proprio sensing, plus direct force feedback to users, without needing a robot👉dex-op.github.io

Check out Any2Track! Really impressive work on whole-body tracking!

Visual manipulation is really challenging for humanoids, and it is impressive to see such interesting results with a depth policy!

Humanoid motion tracking performance is greatly determined by retargeting quality! Introducing 𝗢𝗺𝗻𝗶𝗥𝗲𝘁𝗮𝗿𝗴𝗲𝘁🎯, generating high-quality interaction-preserving data from human motions for learning complex humanoid skills with 𝗺𝗶𝗻𝗶𝗺𝗮𝗹 RL: - 5 rewards, - 4 DR

ResMimic: learns a whole-body loco-manipulation policy on top of general motion tracking a policy Key ideas: (i) pre-train general motion tracking (ii) post-train task-specific residual policy with: (a) object tracking reward (b) contact reward (c) virtual object force

Thrilled to share our work AlphaOne🔥 at EMNLP 2025 2025, Junyu Zhang and I will be presenting this work online, and please feel free to join and talk to us!!! 📆Date: 8:00-9:00, Nov 7, Friday (Beijing Standard Time, UTC+8) 📺Session: Gather Session 4