CV

Download English CV (PDF) · 下载中文简历 (PDF)

Education

• The University of Hong Kong · Aug 2023–Present
  • PhD in Electrical and Computer Engineering · GPA: 3.9 / 4.3
  • Advisors: Dr. Yanchao Yang and Prof. Kaibin Huang
• University of Science and Technology of China · Sep 2019–Jun 2023
  • BSc in Mathematics and Applied Mathematics · School of the Gifted Young
  • Major GPA: 3.87 / 4.3

Research areas

• Tabular foundation model pretraining
• Nonlinear correlation modelling
• LLM reasoning

Research & Industry Experience

Minghong Investment — Machine Learning QR Intern

Aug–Oct 2026

Investigated whether dependence measures can serve as training losses. Using a fixed MLP architecture for five-day A-share return prediction, designed a post-training refitting strategy for dependence-based training, improving mean cross-sectional test IC from 0.105 to 0.12 versus an MSE baseline (both 10-seed ensembles).

Huawei HPC Group (RAFT) — External On-site Student Researcher

Sep 2024–Jan 2026

Developed representation-based diagnostics to monitor model training and validated them on NanoGPT, ViT and Pythia; the resulting research was submitted to ICLR.

ChengQi Asset Management — Quant Research Intern

Aug–Nov 2022

Researched alpha strategies using diverse data sources, with five accepted into the firm’s alpha pool.

Selected Research & Publications

1. InfoNet: Neural Estimation of Mutual Information without Test-Time Optimization

   Z. Hu et al. · ICML 2024 (Oral)

   Developed a tabular foundation model for estimating dependence between random variables, achieving 1,000× faster estimation and higher accuracy on low-dimensional benchmarks. Downstream applications include feature selection, image segmentation and keyframe extraction.

2. InfoAtlas: A Foundation Model for Zero-Shot Statistical Dependence Estimate

   Z. Hu et al. · ICML 2026

   Scaled InfoNet to high-dimensional dependence estimation through richer synthetic data and architecture design. Built data-generation and distributed pretraining pipelines for a 600M-parameter model on 32 H200 GPUs.

3. From Anisotropy to Anomaly: An Online Geometric Diagnostics Framework for Transformer Training

   Z. Hu et al. · ICLR (under review)

   Developed online diagnostics based on statistical dependence among hidden representations to detect anomalies during Transformer training, including overfitting, data anomalies and loss spikes.

4. DensityFM: Foundation-Style Pretraining for Density Estimation

   Z. Hu et al. · ICLR (under review)

   Pretrained a tabular foundation model for density estimation, outperforming all evaluated classical baselines on synthetic and real-world benchmarks. Applications include tail-risk estimation and outlier detection.

Additional Experience

Optiver Quant Focus — Three-day programme for PhD students.

Sep 2026

Team placed 1st in the Quant Challenge; independently solved 5 of the 12 mathematics and brainteaser problems.

Susquehanna (SIG) PhD Brainteaser Battle

Oct 2026

Team placed 2nd; independently solved the competition’s hardest puzzle.

Skills & Tests

• Technical skills: Python · PyTorch · pandas
• Tests: TOEFL iBT 105 · GRE 322 (Quant 170)