I am a master's student in Data Science and Engineering at Zhejiang University (Sept. 2024 - Jul. 2027 expected), advised by Prof. Wei Chen at ZJUVAI within the State Key Laboratory of CAD&CG. Before joining Zhejiang University, I received my B.Eng. degree in Computer Science from Jinan University. My research focuses on LLM-based autonomous agents, especially multi-agent collaboration, autonomous evolution, latent-space communication and reasoning, autonomous evaluation, and reasoning-oriented post-training.
I am currently a Talent Program intern at ByteDance, working on LLM post-training, preference alignment, interactive evaluation, and high-fidelity human-like agent simulation for SFT and RLHF/DPO data generation. Previously, I worked as a research intern at Alibaba Group's Future Living Lab and as a research assistant at THUNLP, Tsinghua University, working with Prof. Zhiyuan Liu, where I contributed to the ChatDev project.
Prospective PhD applications: I am actively seeking Ph.D. positions starting in Fall 2027. I am especially interested in labs working on foundation models, LLM agents, multi-agent systems, autonomous evolution and evaluation, latent-space communication and reasoning, and post-training. Please feel free to reach out if my background may align with your group.
Prospective research internships and collaborations: I am looking for research internship or collaboration opportunities around LLM agents, multi-agent systems, and latent-space communication.
Research Interests
- Autonomous evolution in multi-agent systems: enabling agents to move beyond hand-designed workflows and discover self-improving coordination paradigms.
- Latent-space communication and reasoning: replacing the autoregressive token bottleneck with efficient, high-dimensional machine-native interaction.
- Autonomous evaluation: studying how agents can build self-referential evaluation frameworks in open-ended evolution without human-curated reward models.
News
Selected Publications
* denotes equal contribution.
Experience
- Built the team's first data-flywheel rollout system for multi-turn, task-oriented dialogue, supporting downstream multi-turn dialogue RL and agentic RL experiments.
- Designed self-play simulation and evaluation environments with high-fidelity, persona-conditioned user-simulation agents.
- Worked on LLM post-training, preference alignment, and autonomous agent-driven evaluation for dialogue policies.
- Proposed Interlat (code), a latent-space communication framework that replaces natural-language message passing with temporally aligned hidden-state exchange.
- Designed a token-latent curriculum and learned compression scheme, achieving up to 24x communication speedup at comparable task performance.
- Gained hands-on experience with distributed training on 128+ GPUs; co-authored Online-PVLM, MemCoRL, and Learning to Evolve.
- Led EvoPatient (code), an open-source doctor-patient coevolution system for virtual standardized patients, accepted to ACL 2025.
- Drove the project end-to-end, including research formulation, system implementation, medical-case curation, expert evaluation, and paper writing.
- Improved patient-answer ability from 0.763 to 0.860 after 200 training cases.
- Worked with Prof. Zhiyuan Liu and Prof. Chen Qian on ChatDev-based multi-agent collaboration, studying role organization, information exchange, and coordination.
- Led Croto (code), designing cross-team orchestration with parallel teams, hierarchy partitioning, and greedy aggregation for multi-agent software development.
- Contributed to MACNet (code) and iAgent (code); related OpenBMB/ChatDev projects have accumulated over 33k GitHub stars.
Awards and Honors
- 2025: National Scholarship; Outstanding Graduate Student Scholarship.
- 2024: First-Class Scholarship for Outstanding Graduates.
- Undergraduate: three undergraduate scholarships.
Academic Service
- Reviewer: NeurIPS 2025, ICLR 2026.
Miscellaneous
Beyond research, I enjoy soccer, fencing, snowboarding, billiards, ballroom dancing, piano, photography, and physics. I occasionally write notes on my personal blog.