Chengyu Du (杜承豫)

RESEARCH INTEREST

Chengyu Du (杜承豫) is a second-year M.S. candidate at Fudan University CS. His current research interests are mostly around building Interleaved Agent and World-Play Agent:

  • Interleaved Agent: Enhance the model's capability in handling long-horizon tasks for complex agents and the generalization performance of agents, ultimately achieving agent-in-the-loop self-iteration.
  • World-Play Agent: Developing expansive, autonomously operating open-world World-play. This extends towards the exploration of its interactions with personalized characters and real environments.

Currently, I'm a Research Intern at ByteDance Seed LLM Post-training Team, mentored by Siyu Yuan and Yujia Qin, working on Coding Agent.

🎯 Research Focus: Building Interleaved Coding Agent and World-Play Agent

Email  /  📄 CV  /  Google Scholar  /  Github

profile photo

Experience
  • ByteDance Seed ByteDance Seed - Research Intern (Jan 2026 - Present)
    LLM Post-training Team (Seed-LLM-Horizon), Shanghai, China
    Mentor: Siyu Yuan, Yujia Qin
    Focus: Coding Agent. Participated in Seed 2.0.
  • MiniMax MiniMax - Research Intern (Jan 2025 - Jan 2026)
    LLM Post-training Team, Shanghai, China
    Mentor: Zishan Huang, Pengyu Zhao
    Focus: Talkie role-play Agent Models. Training personalized RM and Role-play Model through case rubrics genRM for RL. Training agents with strong emotional intelligence and excellent tool use. ↑17% total duration of stay, ↑24% conversations during experiments.
  • Alibaba Alibaba Inc. - Research Intern (Jan 2024 - Jan 2025)
    Hangzhou, China
    Mentor: Zulong Chen
    Focus: Progressive Self-reflection in LLMs & Directed Persona Refinement Through RL.

Skills & Fun Facts

🌐 IELTS 7  |  🧠 ENFJ  |  🤖 Skilled at Agent Anything  |  💪 Workaholic  |  💬 Communicating  |  🎉 Organizing Events

News

Selected Publications


(*indicates equal contribution)
MiniMax-M2-her: How we built a Role-Play Agent for the production usage.
MiniMax Team
Technical Report, 2025.
Paper
MiniMax-M2.1: Enhanced Multi-Language Programming and Real-World Complex Tasks.
MiniMax Team
Technical Report, 2025.
Paper
HER: Human Emulation Reasoning for Thinking-in-Character Role-play
Chengyu Du, Xintao Wang, Aili Chen, Weiyuan Li, Rui Xu, Junteng Liu, Zishan Huang, Rong Tian, Zijun Sun, Yuhao Li, Liheng Feng, Deming Ding, Pengyu Zhao, Yanghua Xiao
Under review, ACL 2026.
Paper
MiniMax-M2: A Model Built for Max Coding & Agentic Workflows
MiniMax Team
Technical Report, 2025.
Paper
DIVE: Scaling Diversity in Agentic Task Synthesis for Generalizable Tool Use
Aili Chen, Chi Zhang, Junteng Liu, Jiangjie Chen, Chengyu Du, Yunji Li, Ming Zhong, Qin Wang, Zhengmao Zhu, Jiayuan Song, Ke Ji, Junxian He, Pengyu Zhao, Yanghua Xiao
Under review, ICML 2026.
WebExplorer: Explore and Evolve for Training Long-Horizon Web Agents
Junteng Liu, Yunji Li, Chi Zhang, Jingyang Li, Aili Chen, Ke Ji, Weiyu Cheng, wuzijia, Chengyu Du, Qidi Xu, Jiayuan Song, Zhengmao Zhu, Wenhu Chen, Pengyu Zhao, Junxian He.
Under review, ICML 2026.
MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention
MiniMax Team
Technical Report, 2025.
Paper
Think Thrice Before You Act: Progressive Thought Refinement in Large Language Models
Chengyu Du, J Han, Y Ying, A Chen, Q He, H Zhao, S Xia, H Guo, J Liang et al.
ICLR, 2025.   Poster
Deeper Insight into Your User: Directed Persona Refinement for Dynamic Persona Modeling
A Chen, Chengyu Du, J Chen, J Xu, Y Zhang, S Yuan, Z Chen, L Li, Y Xiao.
ACL, 2025.   Poster
Enhancing confidence expression in large language models through learning from past experience
Haixia Han, Tingyun Li, Shisong Chen, Jie Shi, Chengyu Du, Yanghua Xiao, Jiaqing Liang, Xin Lin
EMNLP, 2025.   Poster
Data-Efficient Selection via Grammatical Complexity in Continual Pre-training of Domain-Specific LLMs
Yizhou Ying, Geng Zhang, Cui Danxin, Chengyu Du, Guanglei Yue, Sihang Jiang, Jiaqing Liang, Yifei Fu, Hailin Hu, Yanghua Xiao
EMNLP, 2025.   Poster
CEM: A Data-Efficient Method for Large Language Models to Continue Evolving From Mistakes
H Zhao, J Han, J Shi, Chengyu Du, J Liang, Y Xiao, W Zhou, Z Sun, F Yu.
CIKM, 2025.   Poster

Patent

Education

Selected Awards
  • Provincial Government Scholarships (Top 1%), 2024
  • University First-Class Scholarships (Top 1%), 2024
  • Excellent Graduates of Zhejiang Province (Top 1%), 2024

Academic Service
  • Reviewer for leading conferences: ACL ARR (2025-), ICLR (2025-), and related workshops.
🤖 🎭 🌍