ByteDance
Research Scientist Graduate (Seed Model - Vision) - 2027 Start (PhD)
San Jose, California · Full-time · Posted Oct 28, 2026
- Education
- PhD in Computer Science, Electrical Engineering, Electrical and Computer Engineering, Physics, Mathematics, or a related discipline.
About the job
Develops and scales vision foundation models for image and video generation, improving their capabilities, training, evaluation, and real-world use in multimodal systems.
What you'll do
- Develop vision foundation models and design data pipelines, pre-training strategies, and post-training methods.
- Improve perception, reasoning, and multimodal understanding; optimize model architectures and training efficiency.
- Develop evaluation frameworks and explore real-world applications of vision models in multimodal systems.
What they ask for
- Completing or recently completed a PhD in a listed or related discipline.
- Proficient in C/C++ or Python, with strong coding, data structures, and algorithm skills.
- Experience with computer vision, multimodal learning, or large-scale model training.
- Strong understanding of deep learning architectures and training methodologies.
Nice to have
- Research experience demonstrated through impactful papers or projects.
- Strong problem-solving and communication skills.
Tools
C/C++, Python
AI focus
Research and development of multimodal vision foundation models for image and video, including perception, reasoning, and multimodal understanding.
Model-written summary · the original posting is authoritative