About

Hello, I am Ke Li, a second-year master’s student at the School of Software Technology, Zhejiang University. I expect to graduate in 2028. I am currently working as a part-time intern at ByteDance. My research interests focus on model lightweighting and efficiency, particularly model compression, knowledge distillation, and efficient reasoning algorithms (e.g., Latent Chain-of-Thought).

Publications

InfoQuant: Shaping Activation Distributions for Low-Bit LLM Quantization

Ke Li, Dong An, Xiaoling Zang, Can Ye, Liang Xie, Qibo Qiu, Chen Shen, Xiaofei He, Wenxiao Wang

EMNLP 2026 Main, 2026

First author

arXiv Code

HEAPr: Hessian-based Efficient Atomic Expert Pruning in Output Space

Ke Li, Zheng Yang, Zhongbin Zhou, Feng Xue, Zhonglin Jiang, Wenxiao Wang

ICLR 2026, 2026

First author

Paper arXiv Code OpenReview

Compressing Neural Networks Using Tensor Networks with Exponentially Fewer Variational Parameters

Yong Qing^†, Ke Li^†, Peng-Fei Zhou, Shi-Ju Ran

Intelligent Computing, Vol. 4, Article 0123, 2025

† Equal contribution

Paper arXiv