About the role#
NVIDIA is developing Cosmos world foundation models and generative AI systems for Physical AI, covering robotics, autonomous driving, smart spaces, and embodied agents. The Cosmos platform facilitates multimodal world understanding, simulation, synthetic data generation, and embodied reasoning. This internship focuses on advancing the frontier of Physical AI and world models.
What you'll do#
- Conduct research in generative AI, multimodal foundation models, world models, and embodied AI.
- Develop algorithms for video understanding and generation, action-conditioned simulation, multimodal reasoning, and policy learning.
- Train and evaluate large-scale models using video, image, language, and robotics or autonomous driving data.
- Collaborate with researchers and engineers across AI, robotics, simulation, and graphics teams.
- Publish research at top conferences and transfer innovations into NVIDIA products.
What you'll need#
- Currently pursuing a PhD in Computer Science, Electrical Engineering, Robotics, or related fields.
- Strong background in generative AI, computer vision, multimodal learning, robotics, or reinforcement learning.
- Prior publication record and research experience.
- Strong Python and PyTorch skills.
- Experience with large-scale foundation model training, video models, VLMs, distributed training, or simulation is preferred.
Location & details#
- Location: Santa Clara, California
- Term: Summer 2026
- Modality: On-site
- Employment: Full-time paid internship
- Compensation: 30 USD - 94 USD per hour, based on location, degree, and experience
