Checkout our new NeurIPS paper on using differentiable trajectory optimization for deep RL and IL! Join our poster session at Friday 11-2. Ziyu will present this paper (he is also applying for grad schools this cycle)

Welcome to the world of Genesis!

I added a post to my blog: ​The Forebot and the Buttlebot​ mtmason.com/forebot-and-bu… Here's the question. What happens when you deploy robots in a warehouse, or in a home? What instructions do you give them? It is only a four-minute read. Hope you like it.

[1/7] Teaching dexterous robot hands to perform functional grasps usually needs hours of teleoperation, manual labeling, or pre-scanning object meshes. Not anymore. 🔥We are excited to introduce Web2Grasp that learns functional multi-finger grasps straight from web images of

Introducing RaC: A data collection protocol that boosts data efficiency by 10x compared to some of the best imitation results. Key idea: scale recovery & correction data systematically => policies can reset+retry when acting (consistent self-correct) => better performance. 🧵0/N

🚨Introducing SPOT: Search over Point Cloud Object Transformations. SPOT is a combined learning-and-planning approach that searches in the space of object transformations. Website: planning-from-point-clouds.github.io Paper: arxiv.org/abs/2509.04645 Code: github.com/kallol-saha/SP…

A closed door looks the same whether it pushes or pulls. Two identical-looking boxes might have different center of mass. How should robots act when a single visual observation isn't enough? Introducing HAVE 🤖, our method that reasons about past interactions online! #CORL2025

check out our new CoRL oral paper on geometric red teaming for breaking robot manipulation policies! We find that small subtle geometry deformations of the object lead to large performance drop of pre-trained manipulation policies.

Check out our new CoRL paper FMVP that adapts to arm motions during assistive dressing! Simulation is great for vision-based pretraining but lack realistic force simulation. FMVP learns vision in sim and then uses real world force to modulate the pretrained vision policy.