About

I am an AI Researcher and Engineer currently at Ping An Technology, where I focus on advancing the frontier of Large Language Models. My work is dedicated to optimizing AI efficiency and reliability, ensuring that complex models are both performant and deployable in practice.

I completed my M.Sc. in Data-Centric AI at HKUST (Guangzhou) 和 earned my B.Sc. in Data Science from Zhongnan University of Economics and Law (ZUEL).

AI Agents LoRA & PEFT Model Compression LLM Efficiency

Recent News

  • 📄 Apr 2026 Paper "Astra: Activation-Space Tail-Eigenvector Low-Rank Adaptation of Large Language Models" accepted at ACL 2026 Findings.
  • 🏢 Mar 2026 Started a new position as AI Researcher & Engineer at Ping An Technology.

Selected Publications

Astra
ACL 2026 Findings
Astra: Activation-Space Tail-Eigenvector Low-Rank Adaptation of Large Language Models
Kainan Liu*, Yong Zhang*, Ning Cheng†, Yun Zhu, Yanmeng Wang, Shaojun Wang, Jing Xiao
→
GRASP
EMNLP 2025 Main Conference
GRASP: Replace Redundant Layers with Adaptive Singular Parameters for Efficient Model Compression
Kainan Liu*, Yong Zhang*, Ning Cheng†, Zhitao Li, Shaojun Wang, Jing Xiao
→
STANDS
Nature Communications
Detecting and dissecting anomalous anatomic regions in spatial transcriptomics with STANDS
Kaichen Xu*, Yan Lu, Suyang Hou, Kainan Liu, Yihang Du, Mengqian Huang, Hao Feng, Hao Wu, Xiaobo Sun†
→

* First author    † Corresponding author