PhD Student | HKUST CSE | Relaxed System Lab

fdingad@connect.ust.hk

Github    LinkedIn

I am interested in probabilistic methods and generative models, my PhD research focuses on diffusion language models.

Education

• Ph.D. Student in Computer Science, the Hong Kong University of Science and Technology (HKUST), 2024 -
• M.Eng. in Computer Science, Shanghai Jiao Tong University (SJTU), 2021 - 2024
• B.Eng. in Computer Science, Shanghai Jiao Tong University (SJTU), 2017 - 2021

Selected Publications

(*: co-first, †: correspondence)

• Beyond Masks: Efficient, Flexible Diffusion Language Models via Deletion-Insertion Processes
  F. Ding*, D. Ding*, S. Chen*, K. Wang, P. Xu, Z. Feng, H. Bai, K. Han, Y. Yan, B. Yuan, J. Sun
  International Conference on Learning Representations (ICLR), 2026.
  ⭐️ Score: 8/8/6/6/6, top 1.3% in ~19000 submissions.
• Improving Subgraph Extraction for Graph Invariant Learning via Graph Sinkhorn Attention
  J. Yan, F. Ding†, J. Sun†, Z. Hu, Y. Zhou, L. Zhu
  IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2026.
• C-NTPP: Learning Cluster-Aware Neural Temporal Point Process
  F. Ding, J. Yan, H. Wang
  AAAI Conference on Artificial Intelligence (AAAI), 2023.

Blogs

• Insertion-Based Discrete Diffusion Language Models Explained
  Presented at the STEM PhD Fellowship Networking Event, Huawei Hong Kong Research Center, 2026.
• Discrete Diffusion from the VAE Perspective