2026 · arXiv
Deep Research in Physical Sciences: A Multi-Agent Framework and Comprehensive Benchmark
Yigeng Jiang, Tengchao Yang, …, Shuchen Pu, …, Shufei Zhang.
AI for Science · Multi-agent systems · LLM RLAI for Science · 多智能体系统 · LLM 强化学习
I am a joint Ph.D. student at the University of Science and Technology of China (USTC) and Shanghai AI Laboratory, affiliated with USTC’s Department of Electronic Engineering and Information Science. I am currently a research intern at Shanghai AI Laboratory, advised by Prof. Wanli Ouyang and jointly mentored by Dr. Shufei Zhang.
My current work focuses on multi-agent systems and reinforcement learning for large language models, especially for scientific reasoning. Before my Ph.D., I received a B.E. in Artificial Intelligence from Huazhong University of Science and Technology (HUST), where I worked with Prof. Renzhi Lu on time-series forecasting and Bayesian methods and with Prof. Shan Tan on medical imaging.
Feel free to reach out if you are interested in AI for Science, agentic systems, or LLM post-training.
我目前是中国科学技术大学(USTC)与上海人工智能实验室(Shanghai AI Laboratory)的联合培养博士生,挂靠中国科大电子工程与信息科学系,现于上海人工智能实验室实习。我的博士导师是欧阳万里教授,联培 mentor 是张舒飞博士。
我目前主要研究多智能体系统与面向大语言模型的强化学习,尤其关注其在科学推理中的应用。本科毕业于华中科技大学人工智能与自动化学院;期间在卢仁智老师指导下开展时序预测与贝叶斯方法研究,在谭山老师指导下学习医学图像方向。
如果你对 AI for Science、多智能体系统或 LLM 后训练感兴趣,欢迎与我交流。
2026 · arXiv
Yigeng Jiang, Tengchao Yang, …, Shuchen Pu, …, Shufei Zhang.
2025 · arXiv
Weida Wang, Benteng Chen, Di Zhang, Wanhao Liu, Shuchen Pu, …, Wanli Ouyang, …, Shufei Zhang.
2025 · arXiv
Weida Wang, Dongchen Huang, …, Shuchen Pu, …, Wanli Ouyang, …, Shufei Zhang.
Affiliated with USTC’s Department of Electronic Engineering and Information Science; currently a research intern at Shanghai AI Laboratory. Early doctoral work explored PXRD/EPR spectrum analysis, diffusion models, and diffusion-model reinforcement fine-tuning. Current work focuses on multi-agent systems, LLM reinforcement learning, and AI for Science.
挂靠中国科大电子工程与信息科学系,目前在上海人工智能实验室实习。博士早期工作包括 PXRD/EPR 谱图解析、扩散模型与扩散模型强化训练;目前主要研究多智能体系统、LLM 强化学习与 AI for Science。
School of Artificial Intelligence and Automation. Undergraduate research in time-series forecasting, Gaussian processes, Bayesian optimization, and medical imaging.
人工智能与自动化学院。本科阶段主要进行时序预测、高斯过程、贝叶斯优化和医学图像相关研究。
GPU-accelerated tools for simulating X-ray diffraction patterns.
用于模拟 X 射线衍射图谱的 GPU 加速工具。
View on GitHub →在 GitHub 查看 →Contributor贡献者A curated knowledge map for PXRD-driven crystal identification, generation, decomposition, and refinement.
一个面向 PXRD 晶体识别、生成、分解与精修的知识图谱。
Explore repository →浏览仓库 →