Humans can teleoperate robots to accomplish complex manipulation tasks. Imitation
learning has emerged as a powerful framework that leverages human teleoperated …

Representing robotic manipulation tasks as constraints that associate the robot and the
environment is a promising way to encode desired robot behaviors. However, it remains …

To endow models with greater understanding of physics and motion, it is useful to enable
them to perceive how solid surfaces move and deform in real scenes. This can be formalized …

We seek to learn a generalizable goal-conditioned policy that enables zero-shot robot
manipulation—interacting with unseen objects in novel scenes without test-time adaptation …

We present Im2Flow2Act, a scalable learning framework that enables robots to acquire real-
world manipulation skills without the need of real-world robot training data. The key idea …

We seek to learn a generalizable goal-conditioned policy that enables diverse robot
manipulation—interacting with unseen objects in novel scenes without test-time adaptation …

How can robot manipulation policies generalize to novel tasks involving unseen object types
and new motions? In this paper, we provide a solution in terms of predicting motion …

We introduce Latent Action Pretraining for general Action models (LAPA), an unsupervised
method for pretraining Vision-Language-Action (VLA) models without ground-truth robot …

Humans can learn to manipulate new objects by simply watching others; providing robots
with the ability to learn from such demonstrations would enable a natural interface specifying …

Both text and video data are abundant on the internet and support large-scale self-
supervised learning through next token or frame prediction. However, they have not been …