About me

Hi! I’m a first-year PhD student at Institute of Computing Technology, Chinese Academy of Sciences (ICT, CAS), fortunate to be advised by Prof. Fei Sun. Previously, I obtained my B.E. in Software Engineering from Nankai University. My research interests include Knowledge Editing, NLP, Trustworthy AI, and Recommender System.

Publications

2024

• The Fall of ROME: Understanding the Collapse of LLMs in Model Editing. [Paper]
  Wanli Yang, Fei Sun, Jiajun Tan, Xinyu Ma, Du Su, Dawei Yin, Huawei Shen
  EMNLP 2024 Findings
• 🦋🌪️ The Butterfly Effect of Model Editing: Few Edits Can Trigger Large Language Models Collapse. [Paper] [Code] [Website]
  Wanli Yang, Fei Sun, Xinyu Ma, Xun Liu, Dawei Yin, Xueqi Cheng
  ACL 2024 Findings
• Blinded by Generated Contexts: How Language Models Merge Generated and Retrieved Contexts When Knowledge Conflicts? [Paper]
  Hexiang Tan, Fei Sun, Wanli Yang, Yuanzhuo Wang, Qi Cao, Xueqi Cheng
  ACL 2024

Experience

• Research Intern, Baidu Inc
  11/2023 - present, Baidu Search, with Dr. Xinyu Ma

• Research Intern, Renmin University of China
  12/2022 - 06/2023, AI Box Lab, advised by Prof. Wayne Xin Zhao

• Research Intern, Nankai University
  10/2022 - 02/2023, AIOps@NKU Lab, advised by Prof. Shenglin Zhang

Project

• Contributing to RecBole: A unified, comprehensive and efficient recommendation library (github.com), 2.9k Stars
  The code replication of Directed Acyclic Graph Factorization Machines for CTR Prediction via Knowledge Distillation, WSDM’ 23

• Contributing to RecBole-GNN: Efficient and extensible GNNs enhanced recommender library based on RecBole (github.com), 126 Stars
  The code replication of LightGCL: Simple Yet Effective Graph Contrastive Learning for Recommendation, ICLR’ 23

• Creating NKUSE.ICU: 南开大学软件学院生存手册 & A website for sharing information about the College of Software at Nankai University, 19 Stars & 4 Forks

Coursework

• Parallel Programming, taught by prof. Gang Wang [Course Project]
  Aiming at the parallel optimization of Inverted Index Intersection utilizing SIMD, Pthread, OpenMP, MPI, and CUDA