- [06/2026] One paper is accepted to ECCV 2026.
- [05/2026] Three papers are accepted to ICML 2026. See you in Seoul!
- [05/2026] I started my internship at
Qwen, Tongyi Lab. - [01/2026] Our paper MMedAgent-RL: Optimizing Multi-Agent Collaboration for Multimodal Medical Reasoning is accepted to ICLR 2026.
- [09/2025] Our paper AceSearcher: Bootstrapping Reasoning and Search for LLMs via Reinforced Self-Play is accepted to NeurIPS 2025 as a Spotlight (top 3.2%) paper!
- [05/2025] Our paper Mitigating Heterogeneous Token Overfitting in LLM Knowledge Editing is accepted to ICML 2025.
Zihan Dong
I am a Ph.D. candidate in Statistics at Rutgers University. Prior to Rutgers, I obtained my B.S. in Mathematics from the School of the Gifted Young, University of Science and Technology of China (USTC), 2023.
Currently, I am working on building long-horizon LLM agents for coding and reasoning through mid- and post-training. Previously, my research interests lay at the intersection of AI and statistics, including semiparametric methods for LLMs, statistical foundations of large language models, and representation learning.
Here is my Google Scholar page and CV.
Currently, I am working on building long-horizon LLM agents for coding and reasoning through mid- and post-training. Previously, my research interests lay at the intersection of AI and statistics, including semiparametric methods for LLMs, statistical foundations of large language models, and representation learning.
- Mid- and post-training for coding agents.
- Statistical Foundations of Large Language Models.
Here is my Google Scholar page and CV.
Open to research discussions, feel free to reach out!