About
About
Agentic LLMs, multimodal reasoning, graph learning, and evaluation-driven AI systems.
Bio
I currently work on large language model systems in Beijing, with a focus on agents that can reason over multi-step tasks, use tools, and improve through evaluation and reinforcement learning feedback loops.
Previously, I completed my master’s research at the Institute of Computing Technology, Chinese Academy of Sciences. My research covered graph anomaly detection, graph out-of-distribution detection, and knowledge reasoning, with publications at AAAI, ACM MM, ICASSP, and related venues.
I received my bachelor’s degree in Computer Science and Technology from Harbin Institute of Technology. I am drawn to problems where research ideas can be turned into reproducible methods, measurable systems, and eventually useful products.
Experience
ByteDance. LLM algorithms, conversational agents, online evaluation, and RL training systems.
Alibaba Token Hub (ATH). Multimodal large models and structured reasoning.
Tencent ARC Lab. Large models, document understanding, and high-resolution multimodal modeling.
Links
I am open to research conversations and collaborations around agentic RL, LLM agents, evaluation, tool use, and interactive learning environments.