Join Us

Research directions

We pursue the four pillars of REAL: Reasoning, Embodied, Agentic, Learnable AI systems.

• Reasoning: reinforcement learning for reasoning, reward modeling, efficient and multimodal reasoning, so that large models can think and decide as deeply as (or beyond) humans.
• Embodied AI: spatial intelligence, embodied reasoning, and world models, closing the perception, reasoning, and action loop in physical and simulated environments.
• Agentic AI: autonomous planning, collaboration, memory, and self-evolution; GUI agents, code agents, tool agents, embodied agents, and multi-agent systems that accomplish complex real-world tasks.
• Learnable AI: models that learn from exploration, feedback, imitation and reflection, and continually self-evolve to new tasks and environments.

What we offer

• Master's and PhD admission quotas in both the College of Computer Science and Technology and the School of Software Technology.
• Year-round undergraduate research internships, with systematic training in research skills and academic innovation.
• Short-term academic visits and collaborations for outstanding master's / PhD students from other institutions.
• Active partnerships with Alibaba, Tencent, and other leading companies, providing rich industry internship opportunities.

What we look for

• Strong execution, clear communication, solid coding ability, and good English reading and writing.
• Genuine interest in large models / AI research and self-driven learning at the frontier.
• Willingness to put in the hard work research demands, and a drive toward breakthrough results.
• Prior research / project experience, LLM background, or co-first-author high-impact publications are a plus.