Papers

Publications, preprints, and work in preparation. See also my CV and Google Scholar.

^* Equal contribution.

1. Preprint

   The Noise-Transmission Burden of Diffusion Network: Why Image Prediction Can Outperform Noise Prediction

   Gyeonghun Kang^*, Pengxi Liu^*, Bingshuo Qian^*, Xiang Cheng

   Preprint · arXiv:2610.34834, 2026

   arXivCode
2. Preprint

   How Do DiTs Denoise? An Aggregator–Predictor View of Attention and MLPs

   Gyeonghun Kang, William Xingxu Chen, Xiang Cheng

   Preprint · arXiv:2610.29854, 2026

   arXivCode
3. ICML26

   Transformers Can Learn Posterior Predictive Distributions In-Context

   Gyeonghun Kang, Changwoo J. Lee, Xiang Cheng

   International Conference on Machine Learning (ICML), 2026

   PaperCode
4. In preparation

   Bayesian Neural Operators for Uncertainty-Aware Predictions of Expensive Partial Differential Equations

   Gyeonghun Kang, Simon Mak

   In preparation

5. Preprint

   Multiscale Cochran–Mantel–Haenszel Scanning for Conditional Dependency

   Gyeonghun Kang, Jialiang Mao, Li Ma

   Preprint · arXiv:2604.19177, 2026

   arXivCode
6. BA26

   Model Selection-Based Estimation for Generalized Additive Models Using Mixtures of g-Priors

   Gyeonghun Kang, Seonghyun Jeong

   Bayesian Analysis, 2026

   First published online in 2024.

   PaperCode