Boxi Yu
Shenzhen
My name is Boxi Yu. I am currently a PhD candidate at The Chinese University of Hong Kong, Shenzhen, and fortunate to be advised by Prof. Pinjia He.
I am interested in both AI Robustness (e.g., developing automated tools for testing and improving the AI software, and developing new testing architectures), and AIOps/MLOps (developing and managing the distributed systems and AI systems).
news
May 20, 2024 | Our extended abstract “DSPy Guardrails: Building Safe LLM Applications via Self-Refining Language Model Pipelines” was accepted by Compound AI Systems Workshop (June 13th, 2024 in San Francisco at Data + AI Summit). |
---|---|
Dec 15, 2023 | Our paper “Testing Graph Database Systems via Equivalent Query Rewriting” was accepted by ICSE’2024. |
Oct 11, 2023 | We introduce “Retromorphic Testing,” a new, general methodology to the test oracle problem. It is a black-box technique, which constructs a dual program architecture to test the target software, inspired by the concept of inverse function. Read the paper |
Sep 2, 2023 | Our paper “Deep Learning or Classical Machine Learning? An Empirical Study on Log-Based Anomaly Detection” was accepted by ICSE’2024. |
May 5, 2023 | Our paper “Automated Testing and Improvement of Named Entity Recognition Systems” was accepted by ESEC/FSE’2023. |
Selected publications
-
ICSETesting Graph Database Systems via Equivalent Query RewritingICSE’24: International Conference on Software Engineering, 2024