I am a PhD student at Peking University, interested in building practical and trustworthy machine learning systems. My current work explores the intersection of federated learning, graph representation learning, and privacy-preserving computation.
北京大学在读博士生，关注联邦学习、图表示学习与隐私保护计算，喜欢把研究想法变成清晰、可复现的代码。
| Area | What I care about |
|---|---|
| Federated Learning | Collaborative learning across distributed and heterogeneous data |
| Graph Representation Learning | Scalable network embedding and graph-based learning |
| Privacy-Preserving ML | Useful learning systems with responsible data boundaries |
| Natural Language Processing | Representation learning for Chinese language understanding |
An experimental implementation of federated network embedding with sparse matrix factorization, including multi-party data partitioning and evaluation utilities.
Python · PyTorch · NetworkX · Graph Learning
An implementation exploring singular value decomposition in a federated setting.
Python · Federated Learning · Matrix Factorization
A Chinese idiom cloze project built on RoBERTa, comparing classification and contrastive-learning approaches.
PyTorch · BERT · Contrastive Learning · Chinese NLP
- 🔬 Exploring efficient and privacy-aware learning on distributed graph data
- 📚 Reading about trustworthy machine learning and scalable representation learning
- 🧭 Documenting research ideas, reproducibility notes, and open-source work
Research is the art of making better questions executable.