Runze Wu (吴润泽)
I am Runze Wu (吴润泽), Head of Collective Intelligence at NetEase Fuxi AI Lab, based in Hangzhou, China. I lead research and product development on Game AI Agents and Human-Robot/Human-AI Collaboration, and drive both toB and toC innovative business. Most notably, I led the creation of China’s first AI-native game with commercialization potential from zero to one — an end-to-end AI-Agent-driven title launched on Steam. I received my Ph.D. from the University of Science and Technology of China (USTC) under Prof. Enhong Chen and Prof. Qi Liu, and completed postdoctoral research at Zhejiang University under Prof. Yunjun Gao.
🎮 Featured work — China’s first AI-native game with commercialization potential. As project lead (team building, design, R&D, marketing & publishing), I built China’s first AI-native game with commercialization potential, fully driven by AI Agents. It launched on Steam and achieved 80% next-day retention, pioneering a new paradigm where generative AI is the core of gameplay rather than an add-on.
My work spans Game AI Agents, Game User Profiling, Trustworthy AI / Human-AI Collaboration, and Educational Data Mining. I have published 80+ papers (~3000 citations) at venues such as KDD, WWW, AAAI, IJCAI, ICDE, CIKM and ACM MM. I also lead NetEase Fuxi’s University Relations (UR) program, collaborating with top universities and institutions including CCF and CAAI.
Research Interests
- Game AI Agents — AI-native games, intelligent NPCs, AI-assisted game development
- Human-AI / Human-Robot Collaboration — crowdsourcing, VLM-based visual inspection, trustworthy AI
- Game User Profiling — LTV/churn prediction, matchmaking, bot detection, attribution & causal analysis
- Educational Data Mining — cognitive diagnosis & proficiency modeling
See my full CV and selected Publications.
关于我 (中文)
我是吴润泽,网易伏羲实验室群体智能负责人,常驻杭州。目前负责游戏 AI Agent 与机器人人机协作方向的技术研发与业务落地,并牵头 toB / toC 创新业务拓展。其中,我主导从零到一研发了国内首款具备商业化潜力的 AI 原生游戏——一款由 AI Agent 端到端驱动、并已上架 Steam 的产品。本科与博士毕业于中国科学技术大学(导师陈恩红教授、刘淇教授),并在浙江大学从事博士后研究(导师高云君教授)。
🎮 亮点工作 —— 国内首款具备商业化潜力的 AI 原生游戏。 作为项目负责人(团队组建、方案设计、技术研发、营销发行全链路),我从零到一打造了国内首款具备商业化潜力、完全由 AI Agent 驱动的 AI 原生游戏。产品上架 Steam,次日留存达 80%,开创了”生成式 AI 作为玩法内核”而非附加功能的全新范式。
研究方向涵盖游戏 AI Agent、游戏用户画像、可信 AI / 人机协作、教育数据挖掘。已发表论文 80+ 篇,引用近 3000 次,成果发表于 KDD、WWW、AAAI、IJCAI、ICDE、CIKM、ACM MM 等会议与期刊。同时负责网易伏羲产学研合作(University Relations),与 CCF、CAAI 等顶尖高校与机构开展合作。
研究方向
- 游戏 AI Agent —— AI 原生游戏、智能 NPC、AI 赋能研发提效
- 人机协作 / 可信 AI —— 众包标注、VLM 视觉审核、模型可解释
- 游戏用户画像 —— LTV / 流失预测、匹配机制、外挂检测、归因与因果分析
- 教育数据挖掘 —— 认知诊断与能力水平建模