About the role#
We are seeking a research intern to develop data-efficient, general algorithms for robot policies. You will work with multiple sensing modalities, including proprioception, images, force, audio, and dense tactile sensing. The team builds Large Behavior Models to create general-purpose robots capable of dexterous tasks in unstructured environments like homes.
What you'll need#
- Strong hands-on experience with machine learning, large datasets, and generative models.
- Research background in world models, action-conditioned prediction, or learned control.
- Proficiency in Python and modern deep learning frameworks like PyTorch.
- Comfort with rapid prototyping and testing in simulation and on physical hardware.
- Interest in contact-rich manipulation and how actions shape physical interactions.
- Preferred experience includes tactile sensors, force-torque sensing, contact simulation, or force-controlled manipulation.
- Currently pursuing a degree in Computer Science, Computer Engineering, Electrical Engineering, Mechanical Engineering, or Data Science.
What you'll do#
- Create working code and hardware prototypes.
- Run experiments using simulated and physical robots.
- Work with large static datasets and a growing corpus of robot data.
- Collaborate frequently with team members.
- Contribute to publishing research in peer-reviewed venues.
Location & details#
- Location: Cambridge, Massachusetts.
- Term: Summer 2027.
- Modality: Hybrid.
- Employment: Full-time, paid internship.


