Role Overview
Meta seeks Research Interns for its Superintelligence Lab, focusing on Agentic AI, Muse, and personal superintelligence. Interns advance agentic LLMs, tool use, personalized agents, post-training, and reasoning/coding at scale. Internships last 12–24 weeks with flexible start dates.
Responsibilities
- Develop novel agentic AI algorithms using machine learning and reinforcement learning.
- Research agentic LLMs, RL environments, and LLM post-training.
- Improve efficiency, scalability, and stability of agentic systems.
- Advance intelligent agents capable of reasoning, tool use, and personalized interaction.
- Collaborate cross-functionally; communicate research plans and results.
- Publish findings, present at conferences, contribute to open source.
- Apply research to Meta product development when applicable.
Requirements
- Work authorization in country of employment.
- Experience with Python.
- Experience building ML/deep learning systems.
- Preferred: publications at top venues (NeurIPS, ICML, ICLR, etc.).
- Preferred: agentic AI contributions, open-source frameworks, deep learning frameworks.
- Preferred: quantitative problem solving, ML experiment design, large-scale data analysis.