Reinforcement Learning AI Research Scientist

Meta•New York, NY

About The Position

Meta's Fundamental AI Research lab is seeking a Research Scientist to drive foundational research aimed at advancing physical AI capabilities. We seek to generate advanced engineer designs such as robotic hardware, mobile vehicles, and novel semiconductors. This role will involve heavy software research engineering, such large-scale data manipulation and simulator integration.

Requirements

Currently has, or is in the process of obtaining a Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience. Degree must be completed prior to joining Meta
Currently has or is in the process of obtaining a PhD degree in Artificial Intelligence, Computer Vision (3D), Physical AI, Machine Learning, relevant technical field, or equivalent practical experience. Degree must be completed prior to joining Meta
Research experience in at least one of the following research areas: reinforcement learning, representation learning, self-supervised learning, multimodal learning, robotics policy development, computer vision (3D), egocentric perception, embodied AI and/or LLMs, control theory, optimization algorithms
Experience in C/C++ and Python and deep learning frameworks (e.g., PyTorch, TensorFlow)
Must obtain work authorization in country of employment at the time of hire, and maintain ongoing work authorization during employment

Nice To Haves

Proven track record of achieving significant results as demonstrated by grants, fellowships, patents, as well as first-authored publications at leading workshops or conferences such as NeurIPS, ICML, ICLR, AAAI, JMLR and Computer Vision (CVPR, ICCV, ECCV, TPAMI)
This position will require knowledge of post-training for LLMs using reinforcement learning techniques. It will also involve novel modalities such as 3D and engineering domain-specific simulators, so computer vision expertise in 3D is welcome as well
Experience integrating and debugging prototype/scientific software-hardware systems including mechanical, aerospace, or civil engineering domain-specific simulation
Experience working and communicating cross-functionally in a team environment
Prior work experience in the fields of mechanical, aerospace, civil engineering or other engineering domains

Responsibilities

Explore and develop novel post-training paradigms for LLMs using reinforcement learning
Explore and develop novel LLM post-training recipes using 3D data
Integrate large-scale simulation into LLM post-training
Explore mechanical, aerospace, civil, and other engineering disciplines and how to enable LLMs to solve key problems in these domains