Research Scientist Intern, Embodied Foundation Models (Evaluation)

In this role, you will work on foundation models for embodied AI, including large-scale pretraining and evaluation. You will train and evaluate multimodal data efficiently in a multi-node distributed system, using PyTorch and Python. You will focus on improving reasoning capabilities and leverage language to enhance embodied agents. You will evaluate performance on open and closed datasets/benchmarks and contribute to rigorous experiments. You will lead high-impact research work and publish at top-tier conferences (e.g., CVPR, NeurIPS, ICCV, CoRL, ICRA). You should have experience in vision-language models, LLMs, and distributed training, and be able to collaborate across cross-functional teams.