Research Scientist in Vision Foundation Model - Seed - Graduates - 2027 Start (BS/MS)

Join the Seed Vision team working on vision foundation models for visual generation and multimodal generative AI. You’ll develop and scale models across image and video, design data pipelines and pre-/post-training strategies, and improve core capabilities like perception, reasoning, and multimodal understanding. The role also focuses on optimizing architectures and training efficiency and exploring real-world applications of vision models within multimodal systems.