Zhan Ling

I am currently a Research Scientist at ByteDance Seed-LLM. I have been working on I have received my Ph.D. at UC San Diego , advised by Prof. Hao Su. Before starting my Ph.D., I received my B.E. in computer science from Yao Class at Tsinghua University.

Research interest

I am an explorer at heart, dedicated to pushing the frontiers of artificial intelligence. My long-term vision is to develop open-world, self-improving intelligent agents that drive us toward super-intelligence and expand the boundaries of human knowledge.

My career is defined by a relentless pursuit of diverse and challenging domains. During my PhD, I ventured broadly across imitation learning, reinforcement learning, robotics, 3D representation learning, and LLM reasoning. After that, I continued this exploration by driving LLM agent post-training at the cutting edge. My work on enhancing LLM tool-use capabilities now powers the Doubao AI Mobile and Doubao Voice Agents. I am continually exploring new technical frontiers, including training long-context LLM agents via reinforcement learning, elevating conversational abilities via multi-agent reinforcement learning, and developing generative reward models for music generation.

Publications and preprints

Papers sorted by recency(*/**/*** = equal contribution). Representative papers are highlighted.

Services