Robotics Control Policy Intern - ML for Manipulation

Meta; Redmond (WA); On-site; USD 55,000 - 83,000; Full time

Job summary

Reality Labs Research in Redmond, WA seeks a Research Scientist Intern at the intersection of robotic control, ML, and human-robot interaction. You will develop data-driven ML-powered control policies for dexterous manipulation and teleoperation across robotics platforms.

Internships range 12–24 weeks with multiple start dates. You will collaborate with world‑leading researchers, publish results, and contribute to state-of-the-art frameworks such as RL, imitation learning, and large behavioral

Qualifications

• Currently pursuing or holding a PhD in Computer Science, AI, Robotics, or a related field.
• Strong understanding of robotic manipulation control policies and advancing the field.
• Experience with ROS, Python, PyTorch or JAX, and related languages.
• Proven ability to work in cross-functional teams and communicate research plans and results.

Responsibilities

• Collaborate on developing control policies for robotics platforms.
• Implement ML frameworks for training policies (LBM, Imitation Learning, RL).
• Develop data collection pipelines for dexterous manipulation and benchmark algorithms.
• Coordinate with researchers and partners to communicate progress and results.

Skills

PhD candidate

Robotics control

Python

PyTorch

JAX

ROS

Cross-functional teamwork

Communication

Reinforcement learning

Education

PhD in CS

Tools

ROS

Python

PyTorch

JAX

Job description

Reality Labs Research in Redmond, WA seeks a Research Scientist Intern at the intersection of robotic control, ML, and human-robot interaction. You will develop data-driven ML-powered control policies for dexterous manipulation and teleoperation across robotics platforms.

Internships range 12–24 weeks with multiple start dates. You will collaborate with world‑leading researchers, publish results, and contribute to state-of-the-art frameworks such as RL, imitation learning, and large behavioral