Research Scientist, Multi-Modal Understanding & Synthesis

What you'll do

• Lead multimodal learning and synthesis research, from problem formulation and data curation through evaluation and real-time prototypes.
• Develop world models and evaluation frameworks for multimodal reasoning, simulation, planning, and prediction.
• Mentor researchers, guide cross-functional technical decisions, and publish findings at peer-reviewed venues.

What they ask for

• Bachelor's degree in Computer Science, Computer Engineering, a relevant technical field, or equivalent practical experience, and a PhD in a related field.
• 6+ years conducting AI research in multimodal learning, generative models, or world models, including leading major initiatives.
• Experience implementing and evaluating multimodal systems with PyTorch or TensorFlow and proficiency in Python.
• Peer-reviewed AI or machine-learning research publications; experience communicating findings and trade-offs to research and engineering audiences.

Nice to have

• Experience developing large-scale multimodal foundation models or vision-language models.
• Experience with world models, predictive learning, or model-based reinforcement learning; first-author publications at NeurIPS, ICLR, or CVPR.

Pay

$184,000/year to $257,000/year USD base salary

Tools

PyTorch, TensorFlow, Python, NeurIPS, ICLR, CVPR

AI focus

Foundational multimodal AI research, including unified understanding and generation across modalities, world models, predictive representations, reasoning, planning, simulation, and multimodal synthesis.