Research Scientist Graduate (Business Integrity) - 2027 Start (PhD)
Core
Research on foundation models and frontier generative technologies, specifically large language models (LLMs), multimodal large models (MLLMs), and agentic AI systems to build capabilities for autonomous decision-making and reflection.
Role type
PhD-level Research Scientist (Large Model/Agentic AI)
Builds
Next-generation content understanding and generation systems for commercialization (advertising and e-commerce content ecosystems)
Domain
Artificial Intelligence, Large Language Models, Multimodal AI, Reinforcement Learning
Deliverable
production ML models
Required skills
Deep theoretical insight in LLM/MLLM architectures, Reinforcement Learning (PPO, GRPO, RLHF), Complex reasoning and long-horizon planning, Diffusion and flow matching algorithms, Algorithm prototyping in PyTorch
Preferred skills
Publications in top AI conferences (NeurIPS, ICLR, ICML, ACL, CVPR, ICCV), Experience with agentic reinforcement learning and multi-agent collaboration
Technologies
PyTorch, LLMs, MLLMs, Diffusion models, Flow matching
Responsibilities
Conduct in-depth research on post-training paradigms for agentic scenarios; Lead algorithmic innovation for agentic AI including long-term memory and knowledge injection; Define and build evaluation systems for next-generation AIGC and agent systems; Track and lead the technical frontier in AI to strengthen team influence
