About the role#
We are looking for a research intern to help advance continual adaptation for robot policies. The goal is to enable robots to improve from experience, personalize their behavior, and build on prior learning to acquire new skills. You will work with robotics and machine learning researchers to develop algorithms and run experiments in simulation and on real robots. This is a 12-week paid internship focused on reliable adaptation with limited robot experience and human effort.
What you'll need#
- Current enrollment in a Ph.D. program in robotics, machine learning, computer science, or a related field.
- Research experience in reinforcement learning, imitation learning, continual learning, or similar methods.
- Strong programming skills in Python.
- Experience with deep learning frameworks like PyTorch or JAX.
- Ability to design experiments, reason critically about results, and work independently with mentorship.
- Clear communication skills.
- Bonus points for experience with generative robot policies, diffusion models, vision-language-action models, online learning, personalization, or modular models.
- Relevant research contributions through publications, open-source projects, or substantial research projects.
What you'll do#
- Work with a mentor to define a research question, develop an approach, and implement it within our research systems.
- Design and run controlled experiments using strong baselines.
- Measure results based on task success, retention of capabilities, adherence to user preferences, and human effort.
- Analyze failure modes and maintain reproducible code and experimental records.
- Collaborate with researchers and engineers to present findings.
- Contribute to technical reports and potential publications in robotics and machine learning venues.
Location & details#
- Location: Cambridge, Massachusetts.
- Term: Summer 2027.
- Modality: Hybrid.
- Employment type: Full-time.
- Pay: $40 to $58 per hour.
- Benefits: Includes medical, dental, and vision insurance, plus paid time off.


