Yiqin Lv 吕怡琴

Ph.D. Candidate @ NUDT | Joint Student @ Tsinghua University

📧 lvyiqin98@nudt.edu.cn 🎓 Google Scholar 💻 GitHub

About Me

I am a fourth-year Ph.D. student in Mathematics at National University of Defense Technology (NUDT), advised by Prof. Zheng Xie and Associate Prof. Qi Wang. I also spent one year as a joint Ph.D. student at the Department of Automation, Tsinghua University under the supervision of Prof. Xiangyang Ji.

I received my B.Sc. degree from Tianjin Normal University in 2020 and M.Sc. degree from National University of Defense Technology in 2022. My research interests lie in Generative AI, Probabilistic Meta-Learning, Natural Language Processing, and Large Language Models.

I am open to collaborations and discussions.

Selected Papers

* indicates equal contribution.

Enhancing Generative Auto-bidding with Offline Reward Evaluation and Policy Search

Zhiyu Mou*, Yiqin Lv*, Miao Xu, Qi Wang, Yixiu Mao, Jinghao Chen, Qichen Ye, Chao Li, Rongquan Bai, Chuan Yu, Jian Xu, Bo Zheng

ICLR 2026 Oral CAAI/CCF-A

[Paper]

A generative auto-bidding framework using offline reward evaluation and policy search to improve advertising performance.

Tail Task Risk Minimization in Meta-Learning from Theoretical Advances to Practical Strategies

Yiqin Lv, Dong Liang, Wumei Du, Zenglin Shi, Zheng Xie, Qi Wang, Meng Wang

IEEE TPAMI 2026 CCF-A Journal IF=18.6

[Paper]

Theoretical and practical advances on tail task risk minimization for more robust meta-learning.

Robust Fast Adaptation from Adversarially Explicit Task Distribution Generation

Qi Wang*, Yiqin Lv*, Yixiu Mao*, Yun Qu, Yi Xu, Xiangyang Ji

KDD 2025 CAAI/CCF-A

[Paper] [Project] [Video]

Adversarially-generated task distributions enable fast and robust meta-learning adaptation.

Theoretical Investigations and Practical Enhancements on Tail Task Risk Minimization in Meta Learning

Yiqin Lv, Qi Wang, Dong Liang, Zheng Xie

NeurIPS 2024 CAAI/CCF-A

[Paper] [Poster] [Slides]

Deeper theoretical analysis and practical improvements for tail risk minimization in meta-learning.

A Simple Yet Effective Strategy to Robustify the Meta Learning Paradigm

Qi Wang*, Yiqin Lv*, Yanghe Feng, Zheng Xie, Jincai Huang

NeurIPS 2023 CAAI/CCF-A

[Paper] [Poster]

A practical and theoretically-supported strategy to make meta-learning more robust.

Multi-scale Graph Pooling Approach with Adaptive Key Subgraph for Graph Representations

Yiqin Lv, Zhiliang Tian, Zheng Xie, Yiping Song

CIKM 2023 CCF-B

[Paper]

Multi-scale graph pooling with adaptive subgraph selection for improved graph-level representations.

View All Papers »

Experience

Alibaba · Taobao & Tmall Group

Research Intern | Mar 2025 – Sep 2025

Intelligent Algorithm Products Division, Ad Algorithm Team (广告算法-全站推广). Focused on generative auto-bidding algorithms. Co-first-authored AIGB-Pearl, accepted at ICLR 2026 Oral.

QiYuan Lab 启元实验室

Research Intern | Jun 2024 – Sep 2024

Intelligent Foundations Research Center. Focused on optimizer design and its applications.

Selected Awards

Outstanding Graduate, NUDT | 国防科技大学校级优秀毕业生 (2022)
NUDT Qiangji First-Class Scholarship | 国防科技大学强基一等奖学金 (2022)
Outstanding Student, Tianjin Municipality | 天津市优秀学生 (2020)
First Prize, National College Mathematics Competition | 全国大学生数学竞赛一等奖 (2019)
National Scholarship | 国家奖学金 (2018)
See CV for full list.

Services

Conference Reviewer

NeurIPS; ICLR; ICML; AAAI; AAMAS; TNNLS.

Invited Talks

RLChina 2024 — Robust Fast Adaptation from Adversarially Explicit Task Distribution Generation
RLChina 2023 — A Simple Yet Effective Strategy to Robustify the Meta Learning Paradigm