CV
A PDF version is available here.
Education
- Ph.D., Electrical and Computer Engineering, The University of Hong Kong, 2023.08 – present
- Advisors: Dr. Yanchao Yang & Prof. Kaibin Huang
- B.Sc., Mathematics and Applied Mathematics, School of the Gifted Young, University of Science and Technology of China (USTC), 2019.09 – 2023.06
- Major-course GPA: 3.87 / 4.3
Research interests
- Foundation models for statistical inference (mutual information, dependency measurement)
- Training dynamics and interpretability of Transformers / LLMs
- Information theory and representation learning
Experience
- Visiting Researcher, Huawei HPC Group (RAFT) — LLM Training Dynamics & Interpretability, 2024.09 – 2026.01
- Designed real-time representation-monitoring metrics for LLM pre-training; validated across multi-scale Transformers for early warning of training anomalies.
- Quantitative Research Intern, Chengqi Asset Management, 2022.08 – 2022.12
- Built, backtested, and shipped 5 alpha factors integrated into the firm’s factor / strategy library.
Selected competitions
- Lingjun Investment Quant Competition — A-Share Market Microstructure Prediction (Kaggle), 2026.03 – 2026.04
- Team ranked Top 13–15. Predicted intraday multi-horizon returns across 500 A-share stocks from Level-2 high-frequency features; benchmarked XGBoost / LightGBM / GRU / LSTM / Transformer with Optuna tuning. Single XGBoost reached test R² = 0.030 (matching the runner-up’s best single model).
Publications
Zhengyang Hu*, Yanzhi Chen*, Hanxiang Ren, Qunsong Zeng, Youyi Zheng, Adrian Weller, Kaibin Huang, Yanchao Yang. (2026). "A Foundation-style Model for Zero-Shot Statistical Dependency Measurement." ICML 2026. (*Equal contribution.)
Zhengyang Hu, Wenyi Fang, Yang Zheng, Yanchao Yang. (2026). "From Anisotropy to Anomaly: Online Geometric Diagnostics during Transformer Training." Under review.
Zhengyang Hu, Song Kang, Qunsong Zeng, Kaibin Huang, Yanchao Yang. (2024). "InfoNet: Neural Estimation of Mutual Information without Test-time Optimization." ICML 2024 (Oral).
Talks
May 20, 2026
Seminar at Department of Electrical and Electronic Engineering, The University of Hong Kong,
July 01, 2024
Conference Oral at International Conference on Machine Learning (ICML),
Skills
- Languages: Python, C/C++, MATLAB, SQL
- ML / DL: PyTorch, distributed multi-GPU training (DeepSpeed / Megatron-style), Transformers, HuggingFace
- Data science: NumPy / Pandas, XGBoost, LightGBM, Optuna
- Tools: Linux, Git, Slurm, Docker
- Languages (Human): Mandarin (native), English (TOEFL iBT 105, GRE 322)
