Theory of Large Language Models (LLMs)
Studying the capabilities of large language models, including but not limited to Transformer-based models, from a theoretical perspective.
Guanyu Cui 崔冠宇
I am currently a PhD student at the Gaoling School of Artificial Intelligence, Renmin University of China, advised by Prof. Zhewei Wei.
Before that, I received my master's degree in Artificial Intelligence from the Gaoling School of Artificial Intelligence, Renmin University of China in June 2025, and my bachelor's degree in Computer Science and Technology from the School of Information, Renmin University of China in June 2022.
Research Interests
My current research focuses on theoretical aspects of artificial intelligence, machine learning, and computation.
Theory of Graph Neural Networks (GNNs)
Studying the capabilities of graph neural networks from a theoretical perspective, with a particular focus on their ability to perform graph algorithms.
Graph Algorithms
Designing and analyzing algorithms for large-scale graphs.
Education
Starting Sep. 2025
PhD Student in Artificial Intelligence
Gaoling School of Artificial Intelligence, Renmin University of China
Sep. 2022 – Jun. 2025
M.Eng. in Artificial Intelligence
Gaoling School of Artificial Intelligence, Renmin University of China
Sep. 2018 – Jun. 2022
B.Eng. in Computer Science and Technology
News
May 1st, 2026
One paper, Position: The Turing-Completeness of Real-World Autoregressive Transformers Relies Heavily on Context Management, was accepted by ICML 2026.
Nov 17th, 2024
One paper, Mixing Time Matters: Accelerating Effective Resistance Estimation via Bidirectional Method, was accepted by KDD 2025.
May 17th, 2023
One paper, MGNN: Graph Neural Networks Inspired by Distance Geometry Problem, was accepted by KDD 2023.
Publications
Authors marked with * are corresponding authors.
Preprint
How Hard Is It for Message-Passing GNNs to Simulate One Weisfeiler-Lehman Color-Refinement Step?
Preprint.
2026
Position: The Turing-Completeness of Real-World Autoregressive Transformers Relies Heavily on Context Management
43rd International Conference on Machine Learning (ICML), 2026.
2025
Mixing Time Matters: Accelerating Effective Resistance Estimation via Bidirectional Method
31st ACM SIGKDD Conference on Knowledge Discovery & Data Mining (KDD), 2025.
2023
MGNN: Graph Neural Networks Inspired by Distance Geometry Problem
29th ACM SIGKDD Conference on Knowledge Discovery & Data Mining (KDD), 2023.
Honors and Awards
Outstanding Graduate of Gaoling School of Artificial Intelligence (Master's)
高瓴人工智能学院优秀毕业生（硕士）
Jun. 2025
Second-Class Graduate Scholarship for Outstanding Academic Performance
研究生学习优秀二等奖学金
Dec. 2023
Outstanding Graduate of Renmin University of China (Undergraduate)
中国人民大学优秀毕业生(本科)
Jun. 2022
Sa-Shixuan Elite Fund Scholarship (First Prize)
萨师煊精英基金奖学金一等奖
Apr. 2022
National Undergraduate Scholarship
本科生国家奖学金
Dec. 2021
National Undergraduate Scholarship
本科生国家奖学金
Dec. 2020
National Undergraduate Scholarship
本科生国家奖学金
Dec. 2019