Beicheng Xu (徐贝澄)

I am a Ph.D. student in Computer Science and Technology at the School of Computer Science, Peking University, advised by Prof. Bin Cui (IEEE Fellow). Before joining Peking University in September 2023, I received my bachelor’s degree in Computer Science and Technology from Wuhan University, where I graduated 1st out of 319 students in my major.

My research focuses on automated research, where I develop intelligent agents to automate literature review, idea generation, experimentation, and scientific writing. A key part of this work is automated machine learning (AutoML): combining large language models with Bayesian optimization to automate feature engineering, algorithm selection, hyperparameter tuning, and model ensembling. I also work on AI for systems, using AI-driven optimization to improve the performance and efficiency of traditional systems such as databases, as well as LLM systems.

I lead AutoSci, an automated research system spanning literature review, ideation, experimentation, and writing, and Openbox, a general-purpose black-box optimization system. My industry collaborations include Tencent, Huawei, and PetroChina.

Contact: beichengxu@stu.pku.edu.cn.

Research Interests

Automated Research and Agents: end-to-end research workflows, from literature review and research ideas to experiments and scientific writing.
AutoML: LLM-guided optimization, feature engineering, combined algorithm selection and hyperparameter optimization (CASH), and model ensembling.
AI for Systems: multi-fidelity optimization, transfer learning, and online configuration tuning under dynamic workloads.

Selected Publications

AutoResearch/AutoML

Tree-Structured Synergy of Large Language Models and Bayesian Optimization for Efficient CASH

Beicheng Xu, Weitong Qian, Lingching Tung, Yupeng Lu, Bin Cui

Conference on Neural Information Processing Systems (NeurIPS), 2026. Poster.

PDF

CoFEH: LLM-driven Feature Engineering Empowered by Collaborative Bayesian Hyperparameter Optimization

Beicheng Xu*, Keyao Ding*, Wei Liu, Yupeng Lu, Bin Cui

ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2026. Oral.

* Equal contribution.

PDF DOI Code arXiv

PSEO: Optimizing Post-Hoc Stacking Ensemble Through Hyperparameter Tuning

Beicheng Xu, Wei Liu, Keyao Ding, Yupeng Lu, Bin Cui

AAAI Conference on Artificial Intelligence (AAAI), 2026. Oral.

PDF Code arXiv

AI for System

MFTune: An Efficient Multi-fidelity Framework for Spark SQL Configuration Tuning

Beicheng Xu*, Lingching Tung*, Yuchen Wang, Yupeng Lu, Bin Cui

International Conference on Very Large Data Bases (VLDB), 2026.

* Equal contribution.

PDF DOI Code

A-Tune-Online: Efficient and QoS-aware Online Configuration Tuning for Dynamic Workloads

Yu Shen*, Beicheng Xu*, Yupeng Lu, Donghui Chen, Huaijun Jiang, Zhipeng Xie, Senbo Fu, Nan Zhang, Yuxin Ren, Ning Jia, Xinwei Hu, Bin Cui

IEEE International Conference on Data Engineering (ICDE), 2025.

* Equal contribution.

PDF DOI Code

All publications

Education

Peking University

Ph.D. student in Computer Science and Technology, School of Computer Science
September 2023 - Present
Advisor: Prof. Bin Cui (IEEE Fellow).

Wuhan University

Bachelor’s degree in Computer Science and Technology, School of Computer Science
September 2019 - June 2023
Rank: 1/319.

Selected Honors

National Scholarship, 2020-2021 and 2021-2022.
CCF Outstanding Undergraduate Award, 2022 (102 recipients nationwide).
BYD Scholarship, Peking University, 2025-2026.
Outstanding Graduate and Outstanding Bachelor’s Thesis, Wuhan University.

Full CV and experience