Fangke Ye

Fangke Ye

Research Scientist



I am a Research Scientis at Meta. Previously, I got my Ph.D. in Computer Science at Georgia Institute of Technology, advised by Prof. Vivek Sarkar in the Habanero Extreme Scale Software Research Laboratory. I received my bachelor’s degree in Computer Science and Technology at Tsinghua University in 2017.

My Ph.D. research focused on integrating machine learning into programming systems to enhance code performance and programming productivity. I developed deep learning methodologies across various domains, including code semantic similarity, type inference, test generation, and compiler optimization. Additionally, I worked on debugging parallel programs using static analysis techniques, such as symbolic execution and polyhedral analysis.

Publications & Preprints

Methods and Apparatus to Determine Refined Context for Software Bug Detection and Correction

Patent No. US 11,782,813 B2

Methods and Apparatus for Automatic Detection of Software Bugs

US Patent Application No. 17/133,238 (Pending)

Methods and Apparatus to Construct Program‑Derived Semantic Graphs

US Patent Application No. 17/133,168 (Pending)

Advanced Graph-Based Deep Learning for Probabilistic Type Inference

Preprint arXiv:2009.05949 [cs.PL]

PDF Cite

MISIM: A Neural Code Semantics Similarity System Using the Context-Aware Semantics Structure

Preprint arXiv:2006.05265 [cs.LG]

PDF Cite Code

Using Polyhedral Analysis to Verify OpenMP Applications are Data Race Free

International Workshop on Software Correctness for HPC Applications (Correctness 2018)

PDF Cite Slides

Detecting MPI Usage Anomalies via Partial Program Symbolic Execution

The International Conference for High Performance Computing, Networking, Storage, and Analysis (SC 2018)

Cite Code Slides


Research Scientist
May 2024 – Present California, USA
Teaching Assistant
Jan 2023 – May 2023 Georgia, USA
  • CS 6245: Parallelizing Compilers
Research Intern
May 2022 – Jan 2023 California, USA
  • Host: Milad Hashemi and Eric Zhang
  • Developed a deep generative model for structured test input using neural module networks and deep reinforcement learning.
Research Intern
May 2019 – Aug 2019, May 2020 – May 2021 California, USA
  • Host: Justin Gottschlich
  • Developed MISIM, a neural code semantic similarity system, using a novel code representation named context‑aware semantics structure and deep metric learning neural networks.
Teaching Assistant
Jan 2019 – May 2019 Georgia, USA
  • CS 4240: Compilers and Interpreters
Research Intern
May 2018 – Aug 2018 California, USA
  • Host: Markus Schordan
  • Developed a polyhedral analysis based approach to verify OpenMP parallel affine loop nests are data race free using the ROSE compiler infrastructure.
Ph.D. in Computer Science
Aug 2017 – May 2024 Georgia, USA
Exchange Student at the Department of Computer Science
Aug 2015 – Dec 2015 Texas, USA
B.Eng. in Computer Science and Technology
Aug 2013 – Jul 2017 Beijing, China


  • First Place in the 6th ISC‑HPCAC Student Cluster Competition, 2017

    • Participated as the leader of the Tsinghua Student Cluster Competition Team.
  • Champion in ASC Student Supercomputer Challenge, 2017

    • Participated as the leader of the Tsinghua Student Cluster Competition Team.
  • Second Place in the 5th ISC‑HPCAC Student Cluster Competition, 2016

    • Participated as a member of the Tsinghua Student Cluster Competition Team.
  • First Prize in the “Challenge Cup” Tsinghua University Student Science and Technology Works Competition, Tsinghua University, 2015

  • Scholarship of Academic Excellence, Tsinghua University, 2014/2015

  • First Prize in the National College Student Physics Contest, China, 2014