Wenkang Han

Wenkang Han

AI Algorithm Engineer at Tencent.

💼 About Me

I received my B.S. from Zhejiang A&F University ZAFU, where I actively participated in ACM competitions and advanced to the 2020 EC-Final. I obtained my master’s degree from Zhejiang UniversityZJU (2023.09-2026.03), advised by Jingyuan Chen. I am currently an AI Algorithm Engineer at Tencent. My research focuses on visual generation and multimodal understanding, with on Google Scholar.

Building on my internship experience with RL-based GUI agents, my current research centers on Multimodal Agents within digital environments, spanning web browsers, OS interfaces (PC/Mobile), and gaming scenarios. I am particularly driven by three core challenges in this field:

  • Long-Horizon Memory: Enabling agents to efficiently store and retrieve action experiences during extended interactions.

  • Agentic Reinforcement Learning: Exploring scalable agent training paradigms such as OpenClaw-RL: Train Any Agent Simply by Talking, which lower the barrier of RL-based agent development through natural language driven training.

  • World Model Grounded VLA: Aligning Vision-Language-Action policies with physical priors from world models, enabling more grounded and generalizable decision-making.

I am open to collaborations in multimodal understanding and reinforcement learning. If you are recruiting researchers or algorithm engineers, please feel free to contact me at dqyzhwk@gmail.com.

📝 Publications

  • ACM-MM 2025 Show and polish: reference-guided identity preservation in face video restoration. Wenkang Han, Wang Lin, Jingyuan Chen, et al.

  • ECML-PKDD 2025 Revisiting applicable and comprehensive knowledge tracing in large-scale data. Yiyun Zhou, Wenkang Han, Jingyuan Chen.

  • IJCAI 2025 Contrastive Cross-Course Knowledge Tracing via Concept Graph Guided Knowledge Transfer. Wenkang Han, Wang Lin, Jingyuan Chen, et al.

  • CVPR 2025 Towards transformer-based aligned generation with self-coherence guidance. Shulei Wang, Wang Lin, Hai Huang, Hanting Wang, Sihang Cai, WenKang Han, Zhou Zhao, et al.

  • NeurIPS 2024 $E^3$: Exploring Embodied Emotion Through A Large-Scale Egocentric Video Dataset. Wang Lin, Wenkang Han, Yueying Feng, Jingyuan Chen, et al.

  • ACL 2024 Mpcoder: Multi-user personalized code generator with explicit and implicit style representation learning. Zhenlong Dai, Chang Yao, WenKang Han, Jingyuan Chen, et al.

🏅 Honors and Awards

  • 2025.10 National Scholarship (master) (Top 3%).
  • 2021.4 The 2020 ICPC Asia-East Continent Final (EC-Final) Contest, 🥉Bronze Medal.

  • 2021.4 The 2020 ICPC Asia Kunming Regional Contest, 🥈Sliver Medal.

  • 2020.11 The 2020 ICPC Asia Nanjing Regional Contest, 🥉Bronze Medal.

👨‍🏫 Academic Service

Conference Reviewer

  • AAAI 2026, ICLR2026

Journal Reviewer

  • TOIS 2026

💼 Experience

  • 2026.03 - Present, AI Algorithm Engineer at Tencent, Shenzhen.

📖 Educations

  • 2023.09 - 2026.03, M.S., Zhejiang University, Hangzhou.
  • 2018.09 - 2022.06, B.S., Zhejiang A&F University, Hangzhou.

👨‍💻 Internships

  • 2025.01 - 2025.08, Meituan M17. Topic: Generalist Multi-modal Agent, EndtoEnd Online RL.