多模态世界模型算法研究员-Seed
Core
Research and develop multimodal world models, focusing on multimodal understanding, generative AI, reinforcement learning, and AIGC to build virtual world agents and optimize large-scale foundation models.
Role type
Senior Research Scientist (Multimodal World Models & AIGC)
Builds
Multimodal agents for GUI/games, scalable foundation models, and virtual/reality environment simulations
Domain
Artificial Intelligence, Multimodal Learning, Generative AI, Reinforcement Learning
Deliverable
production ML models
Required skills
Multimodal understanding, Generative AI, Reinforcement Learning, Computer Vision, Large Language Models, Model Optimization, Data Synthesis, Scalable Oversight, GUI/Game Agent Development
Preferred skills
Publications in CVPR/ECCV/ICCV/NeurIPS/ICLR/SIGGRAPH, ACM/ICPC/NOI/IOI/Kaggle awards, Leadership in high-impact multimodal/foundation model projects
Technologies
C/C++, Python, Pre-training, Simulation, RAG, Visual CoT
Responsibilities
Explore frontier technologies in multimodal understanding and generative AI; Optimize large-scale multimodal foundation models and system performance; Develop multimodal agents for virtual environments; Model virtual and real-world environments using pre-training and simulation techniques.
