Mingwei Zheng
Scholar

Mingwei Zheng

Google Scholar ID: 287g9UYAAAAJ
Purdue University
Large Language ModelsSoftware Engineering
Citations & Impact
All-time
Citations
49
 
H-index
4
 
i10-index
3
 
Publications
10
 
Co-authors
9
list available
Resume (English only)
Academic Achievements
  • Published 'SFA-Miner: Mining Path-Sensitive API Usage Patterns via Symbolic Finite Automata' at IEEE S&P 2026
  • Published 'RFCAudit: AI Agent for Auditing Protocol Implementations Against RFC Specifications' at ASE 2025
  • Co-first author of 'CORE: Benchmarking LLMs’ Code Reasoning Capabilities through Static Analysis Tasks' at NeurIPS 2025 (Datasets and Benchmarks Track), Spotlight (56 out of 497 accepted papers)
  • Contributed to 'IntenTest: Stress Testing for Intent Integrity in API-Calling LLM Agents' at NeurIPS 2025
  • First author of 'Validating Network Protocol Parsers with Traceable RFC Document Interpretation' at ISSTA 2025
  • First author of 'Large Language Models for Validating Network Protocol Parsers' at LangSec 2025
  • First author of 'ParDiff: Practical Static Differential Analysis of Network Protocol Parsers' at OOPSLA 2024, awarded ACM SIGPLAN Distinguished Paper Award (7 out of 148 accepted papers)
  • Co-author of 'Lifting Network Protocol Implementation to Precise Format Specification with Security Applications' at CCS 2023
  • First author of 'Why Do Developers Remove Lambda Expressions in Java?' at ASE 2021
  • Co-author of preprint 'PR²: Peephole Raw Pointer Rewriting with LLMs for Translating C to Safer Rust', arXiv:2505.04852 (2025)
  • Received Spotlight recognition at NeurIPS 2025 (Datasets and Benchmarks Track)
  • Recipient of ACM SIGPLAN Distinguished Paper Award (OOPSLA 2024)
Background
  • Ph.D. candidate in the Department of Computer Science at Purdue University, advised by Prof. Xiangyu Zhang since 2021
  • Research focuses on improving software reliability and efficiency
  • Builds LLM agents or program analysis tools to automate software development tasks such as code generation, software testing, and program repair
  • Aims to enhance software correctness, robustness, and trustworthiness
  • Actively seeking full-time industry roles starting summer 2026