About Me
I’m Anwar Hossain Zahid, a Ph.D. candidate in Computer Science at Iowa State University, advised by Professor Wei Le.
My research lies at the intersection of Machine Learning and Software Engineering, with a focus on debugging and reliability of ML systems. Specifically, I work on developing differential debugging frameworks for machine learning model versions and techniques for numerical stability analysis across platforms and compilers.
Previously, I worked as a Computing Graduate Intern at Lawrence Livermore National Laboratory, where I contributed to testing GPU numerics across heterogeneous architectures (NVIDIA and AMD).
My research has resulted in publications at top-tier venues like ACM FSE and preprints addressing numerical bugs in ML models, cross-GPU numerical inconsistencies, and evaluation of large language models for social good.
Before starting my Ph.D., I accumulated four years of industry experience in software engineering, integration, and R&D, working on projects involving ERP systems, AI-driven solutions, financial technology automation, and mobile platforms.
Research Interests
- ML Reliability & Debugging
- Numerical Analysis & Stability
- Cross-Platform GPU Computing
- AI Safety and Robustness
Let’s Connect
- 🌐 Google Scholar
- 💻 GitHub
- ✉️ Email: ahzahid@iastate.edu
💡 I am open to collaborative research opportunities, industry partnerships, and internships in the areas of ML reliability, debugging, and performance optimization. Feel free to reach out!