Yingyu Lin

Publications and Preprints

Beyond Length: Quantifying Long-Range Information for Long-Context LLM Pretraining Data

Haoran Deng*, Yingyu Lin*,  Zhenghao Lin, Xiao Liu, Yizhou Sun, Yi-An Ma, Yeyun Gong

Preprint

[arxiv]

On the Complexity Theory of Masked Discrete Diffusion: From poly(1/ε) to Nearly ε-Free

Xunpeng Huang*, Yingyu Lin*, Nishant Jain, Kaibo Wang, Difan Zou, Yian Ma, Tong Zhang

Preprint

[arxiv] 

Almost Linear Convergence under Minimal Score Assumptions: Quantized Transition Diffusion

Xunpeng Huang*, Yingyu Lin*, Nikki Lijing Kuang, Hanze Dong, Difan Zou, Yian Ma, Tong Zhang

Preprint

[arxiv] 

Purifying Approximate Differential Privacy with Randomized Post-processing

Yingyu Lin*, Erchi Wang*, Yi-An Ma, Yu-Xiang Wang

Advances of Neural Information Processing (NeurIPS), 2025. (Spotlight)

Theory and Practice of Differential Privacy (TPDP) Workshop, 2025. (Contributed talk)

[arxiv] 

A Skewness-Based Criterion for Addressing Heteroscedastic Noise in Causal Discovery

Yingyu Lin*, Yuxing Huang*, Wenqin Liu*, Haoran Deng*, Ignavier Ng, Kun Zhang, Mingming Gong, Yi-An Ma, Biwei Huang

International Conference on Learning Representations (ICLR), 2025.

[arxiv] 

Tractable MCMC for Private Learning with Pure and Gaussian Differential Privacy

Yingyu Lin*, Yi-An Ma*, Yu-Xiang Wang*, Rachel Redberg, Zhiqi Bu

International Conference on Learning Representations (ICLR), 2024 .

[arxiv] [slides] 

Contact

 Email: yil208@ucsd.edu

Google Scholar Page: Yingyu Lin