About me
I am currently a researcher on system security team at Visa Research, my research area lies in the intersection of Programming Language, Deep Learning and Artificial Intelligence. Previously I obtained my PhD in Computer Science from UC Davis under the supervision of Prof. Zhendong Su.
Visa Research is an emerging industrial research lab. We focus on fundamental research, encourage university collaboration, and emphasize academic publication. If you have a great publication record and are interested in pursuing a research career in our lab. Feel free to reach out at [first name](no space, comma, hyphen, etc.)[last name]@visa.com
Contact
Research Highlights
Learning semantic program embeddings is an important problem. Precise and scalable program representation enables the application of deep neural networks to a wide range of program analysis tasks. My reseach tackles this problem from three different angles: dynamic (i.e. learning from execution/symbolic traces, ICLR'18 and S&P'21), static (i.e. learning from source code, OOPSLA'20 and ICLR'20) and hybrid (i.e. learning from both, PLDI'20).
PhD Students/Post-docs/Research Scientists (co-) Advised
- Guoren Li from University of California, Riverside (06/2022 - 09/2022)
- Jihye Choi from University of Wisconsin-Madison (06/2022 - 09/2022)
-
Dr. Yizhen Wang from University of California, San Diego (06/2020 - Now)
- [NEURIPS 2022] Robust Learning against Relational Adversaries (Oral Paper Presentation)
-
Dr. Yu Wang from Nanjing University (02/2019 - 12/2020)
- [OOPSLA 2020] Learning Semantic Program Embeddings with GINN (Distinguished Paper Award)
- Shitong Zhu from University of California, Riverside (02/2021 - 12/2021)
- Ahmed Abusnaina from University of Central Florida (06/2021 - 09/2021)
- Md Rafiqul Rabin from University of Houston (02/2019 - 2021/02)
-
Ziyang Li from University of Pennsylvania (06/2020 - 12/2020)
- [S&P 2021] ARBITRAR: User-Guided API Misuse Detection
- Elizabeth Dinella from University of Pennsylvania (04/2019 - 07/2020)
- Daniel Yahyazadeh from University of Iowa (06/2019 - 09/2019)
Professional Service
- Program Committe: PLDI 2023, ICSE SRC 2023, PLDI 2021, PLDI 2020, MAPL 2020
- Review Committe: ICLR 2023, NeurIPS 2022, ICML 2022, ICLR 2022, NeurIPS 2021, ICML 2021, NeurIPS 2020, USENIX Security 2019
- Journal Reviewing: TMLR 2022, Empir.Softw.Eng. 2021, TSE 2021, JSS 2021, ASE 2020, J. Symb. Comput. 2017
Past Employment
02/2017-06/2018 Microsoft Research Redmond
Papers: PLDI 2018, ICLR 2018
Patent: Data-driven feedback generator for programming assignments
Products: Microsoft C# course, Microsoft Python Course
Post: AI for Education: Individualized Code Feedback for Students
09/2016-12/2016 Facebook Inc.
03/2016-09/2016 Microsoft Research Redmond.
09/2010-07/2012 Siemens Corporate Technology.
01/2010-06/2010 Siemens Corporate Research.
News
I will be reviewing for ICLR 2022.
I will be reviewing for NeurIPS 2021.
I will be reviewing for ICML 2021.
GINN has won a distinguished paper award.
I will be serving on the program committe for PLDI 2021.
GINN to appear in OOPSLA 2020.
I will be reviewing for NeurIPS 2020.
Blended program embedding to appear in PLDI 2020.
I will be serving on the program committe for MAPL 2020.
I will be serving on the program committe for PLDI 2020.
I will be joining Visa Research.
A new Microsoft Python Course shipped with the feedback technology is deployed on edX.
Search, Align and Repair for Feedback Generation to appear at PLDI 2018.
Dynamic Neural Program Embedding to appear at ICLR 2018.
ETS reached out for possible adoption of RPM generation.
I am leading a second team to adapt the feedback technology to a new Microsoft Python course to be launched after the new year.
The product is officially deployed to integrate with Microsoft C# course on edX. Here is the blog post .
I am leading a team to productionalize the research on feedback generation for MOOC.