Biography

Hello! I am Bo-Han Yang (杨博涵), currently an undergraduate student at School of Artificial Intelligence in Nanjing University. My supervisors are Assistant Researcher Lei Yuan and Professors Yang Yu, and I am also a member of the LAMDA Group led by Professor Zhi-Hua Zhou.

My research interests include reinforcement learning, such as offline reinforcement learning, reinforcement learning training for large language models, and multi-agent reinforcement learning.

Publications

* indicates equal contribution

Learning to Reuse Policies in State Evolvable Environments [paper][code][poster]

Ziqian Zhang*, Bohan Yang*, Lihe Li, Yuqi Bian, Ruiqi Xue, Feng Chen, Yi-Chen Li, Lei Yuan, Yang Yu

The 42nd International Conference on Machine Learning ICML 2025

We addresse the performance degradation of RL policies when state features (e.g., sensor data) evolve unpredictably by proposing Lapse, a method that reuses old policies by combining them with a state reconstruction model for vanished sensors and leverages past policy experience for offline training of new policies.

Generalizable Offline Multi-Agent Reinforcement Learning with Flexible Skill Modulations Retrieval

Lei Yuan, Bohan Yang, Ziqian Zhang, Lihe Li, Yang Yu

Submit to the IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

We propose GOMAS, a transformer-based framework for multi-task offline multi-agent reinforcement learning. It includes a pre-trained encoder for task-agnostic representations and a skill modulation mechanism to handle action diversity. GOMAS is evaluated across multiple benchmarks, showing effective learning and generalization in multi-agent environments.

Awards & Honors

  • 2025.6 Outstanding Graduate of Nanjing University (南京大学优秀毕业生)

  • 2025.5 Special Award of the UbiPoker AI Contest. $5,000 CNY (九坤德州扑克AI对抗赛特色奖)

  • 2024.12 Yang Lanyun Leadership-Oriented Talent Scholarship, $5,000 CNY (杨蓝云领导型人才奖学金)

  • 2024.11 Grand Prize of the Challenge Cup Leaderboard Challenge Special Competition (“挑战杯”揭榜挂帅专项赛特等奖)

  • 2024.5 Grand Prize of the Jiangsu Collegiate Computing Competition (江苏省计算机设计大赛特等奖)

  • 2024.3 Research & Competition Star of the School of Artificial Intelligence, Nanjing University, $2,000 CNY (南京大学人工智能学院科研竞赛之星)

  • 2023.10 Champion of the E Fund Cup AI+ Collegiate Innovation Competition, $10,000 CNY (易方达资产杯“AI+”大学生创新技能挑战赛第一名)

  • 2019.12 First Prize of the National Olympiad in Informatics in Provinces (NOIP 吉林省一等奖)

Miscellaneous

  • I am a fan of Touhou Project doujin music, and I also really enjoy trance, especially uplifting trance. Perhaps we have something in common!

  • I built and deployed the LAMDA RL LAB homepage based on VitePress and Tailwind CSS.

  • I designed the emblem for Kai Jia College of Nanjing University. Kai Jia College is a freshman college for the computer science discipline at Nanjing University.