There are videos of dancing robots everywhere but they are very rarely using perception to decide what to do like this. Just need to integrate manipulation

How do we build long-horizon memory of dynamic environments in existing world models? Excited to share our recent work on flow-equivariant world models, which helps us model dynamic objects in our memory by using group symmeteries!

🤖Meet the OmniHand Pro 2025 from AGIBOT: High DOF: 19 Total DOF (12 active + 7 passive) in a highly compact, lightweight 750g design. Precision Sensing: Features 150+ taxels with 0.01N force resolution. Critical sensitivity for demanding tasks. Industrial Ready: Provides up

Being-H0.5 Scaling Human-Centric Robot Learning for Cross-Embodiment Generalization

Q-learning with adjoint matching is our latest method for offline RL & offline -> online RL with diffusion/flow models, and (so far?) the best-performing method we've developed for training flow policies with RL. Check out Colin's thread!

Booster playing soccer, fully autonomous. Delivering this robot to our client @FastFintex in Texas

RoboMaster is accepted to ICLR 2026🎉 We verify a paradigm for robot learning: Interactive 2D Trajectory Input →I2V Robotic Manipulation Demonstrations → Actions via Inverse Dynamic Model 1. Project Page: fuxiao0719.github.io/projects/robom… 2. Code: github.com/KlingTeam/Robo…

What if your video generator could refine itself—at inference time? ❌No new models. ❌No retraining. ❌No external verifier. 💡 Introducing Self-Refining Video Sampling By reinterpreting a pretrained generator (Wan2.2, Cosmos) as a denoising autoencoder, we enable iterative

DynamicVLA A Vision-Language-Action Model for Dynamic Object Manipulation

x.com/i/article/2018…

x.com/i/article/2021…

Introducing MolmoSpaces, a large-scale, fully open platform + benchmark for embodied AI research. 🤖 230k+ indoor scenes, 130k+ object models, & 42M annotated robotic grasps—all in one ecosystem.

Robot foundation models are limited by costly real data, while simulation data is plentiful but visually mismatched to reality. We present Point Bridge, a method that enables zero-shot sim-to-real transfer for robot learning with minimal visual alignment. pointbridge3d.github.io

If we train VLAs to respond to diverse multimodal prompts, then we can steer them better: [grasp the carrot]/[move to x,y,z]/[put the carrot on the plate]. With many levels of detail, powerful VLMs can step in and steer the model to success much more often! More below 👇

The interplay between various components in RL can sometimes be very frustrating 😅 An experiment that wouldn't work is now fine after turning off obs and value normalization. Every other experiment works better with both turned on.

🚀Excited to share that our paper EgoX👀 has been accepted to #CVPR2026 ! Huge thanks to my co-first authors (taewoongkang, dohyeon), co-authors (Minho Park,junhahyung) and Prof. Jaegul Choo. See you in Denver!🏔️ #VideoGeneration #WorldModeling #Robotics

1/ World models are getting popular in robotics 🤖✨ But there’s a big problem: most are slow and break physical consistency over long horizons. 2/ Today we’re releasing Interactive World Simulator: An action-conditioned world model that supports stable long-horizon interaction.

Fast Foundation Stereo + SAM2 basically = zero shot Foundation Pose😂 No CAD model, no object image, just click the target. Run directly on my 3070 at 13fps, with a 30$ stereo camera (calibrated in 10min) Thanks Bowen Wen for his contribution to the community!