I am Xiaohan Wang, a Ph.D. student in Computer Science at Vanderbilt University, advised by Kevin Leach. My research centers on reliable LLM-assisted decompilation and binary analysis, combining semantic recovery, program analysis, and behavioral feedback to reconstruct readable, executable code from compiled binaries.
As a Graduate Research Assistant at Vanderbilt, I contribute to healthcare cybersecurity and autonomous defense efforts under the ARPA-H DIGIHEALS and ARPA-H UPGRADE programs. Earlier in my Ph.D. and undergraduate studies, I worked on machine learning for scientific visualization and neural field exploration with Prof. Matthew Berger, and held a research internship at The Hong Kong University of Science and Technology.
Education
- Ph.D. in Computer ScienceVanderbilt UniversityAug 2023 – Expected May 2028
- B.Eng. in Software EngineeringNanjing Normal University, Yingcai Honors CollegeSep 2019 – Jun 2023
Research Interests
- Code and program understanding with large language models
- Large-scale scientific simulation data exploration using machine learning techniques
Selected Publications
-
STILL: Recovering Lowered STL Semantics for LLM-assisted C++ Decompilation -
ICL4Decomp: Towards Re-executable Reverse Engineering -
Constraint-Guided Multi-Agent Decompilation for Executable Binary Recovery*Equal contribution. -
Seeing the Many: Exploring Parameter Distributions Conditioned on Features in Surrogates Best Paper Award -
Topology Guidance: Controlling the Outputs of Generative Models via Vector Field Topology -
Who's Pushing the Code? An Exploration of GitHub Impersonation -
EmotionLens: Interactive visual exploration of the circumplex emotion space in literary works via affective word clouds
News
- Sep 2026Our paper “Constraint-Guided Multi-Agent Decompilation for Executable Binary Recovery” has been accepted to NeurIPS 2026 Workshop on Evaluation of Interactive Agents.
- Aug 2026Our paper “STILL: Recovering Lowered STL Semantics for LLM-assisted C++ Decompilation” has been accepted to EMNLP 2026 Main Conference.
- Aug 2026Our paper “ICL4Decomp: Towards Re-executable Reverse Engineering” has been accepted to Findings of EMNLP 2026.
- Nov 2025Our paper “Seeing the Many: Exploring Parameter Distributions Conditioned on Features in Surrogates” is accepted and received the Best Paper Award at the IEEE Workshop on Uncertainty Visualization.
- May 2025Our work “Topology Guidance: Controlling the Outputs of Generative Models via Vector Field Topology” is released as an arXiv preprint.
- Mar 2025Our collaborative project “EmotionLens: Interactive visual exploration of the circumplex emotion space in literary works via affective word clouds” is published in Visual Informatics.
- Mar 2025Our paper “Who's Pushing the Code? An Exploration of GitHub Impersonation” is accepted by ICSE.
- 2025I began a new research direction under the guidance of Prof. Leach.
- Nov 2024I passed my Ph.D. preliminary examination.
- Aug 2023I started my Ph.D. in Computer Science at Vanderbilt University under the supervision of Prof. Berger.
- 2020I was selected for Yingcai Honors College at Nanjing Normal University.
Awards
- 2020 – 2022Outstanding Student Scholarship — First Prize Top 5% of cohort
