AI Research Scientist: Multimodal Foundation Models - Architecture, Pre-Training & Distillation

Tech Stack

Responsibilities

• Advance the architectures, pre-training methods, and distillation techniques for highly capable multimodal models.
• Develop foundation models that combine broad intelligence with strict memory, latency, energy, and privacy requirements for on-device deployment.
• Shape new research directions and conduct ambitious model training experiments at scale.
• Translate successful research ideas into foundation-model technologies that can reach Apple products.
• Contribute to publications and the open sourcing of selected models, research artifacts, or tools where appropriate.

Culture

Requirements