Weijie Xu

About Weijie Xu

Engineer and Researcher in data science and machine learning (LLM). Check out my website: weijiexu.comPublish in COLM, ACL, EMNLP and TMLR as first author in the fields of Responsible AI, LLM Evaluation, Tabular LLM, and Topic Modeling. See full publication list:

Experience

1. Senior Applied Scientist Tech Lead

   Amazon

   Jul 2020 — Present · Seattle, WA, US

   Lead a system that evaluates AI applications both before and after deploymentPre-production testing focuses on responsible AI aspects:FairnessAppropriate refusal capabilitiesPrivacy protectionHallucination detectionCheck out relevant publications:Over-refusal: https://arxiv.org/pdf/25•••••54 COLM 2025Group level Fairness: COLM 2025Post-production monitoring uses fine-tuned LLMs to evaluate real-time performance:Token debiasing: https://arxiv.org/abs/24•••••57 ACL 2025Token debiasing on select all that apply questions: https://arxiv.org/pdf/25•••••43Previous work:LLM in Tabular Prediction. Created and deployed a unified attrition prediction system that is explainable, accurate and fast to train.https://assets.amazon.science/f1/0c/5f8a587c452d9b1c687f70b731ab/large-language-models-llms-on-tabular-data-prediction-generation-and-understanding-a-survey.pdf TMLR 2024Invented and deployed a topic modeling system that is semi-supervised.Topic Modeling Publications:https://aclanthology.org/2023.findings-acl.271/ ACL 2023https://aclanthology.org/2023.findings-emnlp.606/ EMNLP 2023https://arxiv.org/abs/23•••••78https://arxiv.org/abs/23•••••04Created the first internal HR agent prototype for conducting survey.https://aclanthology.org/2024.nlp4hr-1.5/ Differential Privacyhttps://arxiv.org/abs/23•••••61https://arxiv.org/abs/23•••••75

Education

• University of Toronto

  Honors Bachelor of Science with high destinction, Math and its application in stats specialist and Computer Science Minor

• Huazhong University of Science and Technology

  Bachelor's degree, Mathematics and Statistics

• University of California, Berkeley

  Master's degree, Statistics

Skills

• Data Analysis
• R
• Python
• Machine Learning
• Time Series Analysis
• Statistical Computation
• Leadership
• Algorithm
• Data Structure
• Matlab