About

I’m Md Rahinul Islam Bhuiyan (Rahin), a second-year Artificial Intelligence undergraduate at the University of Malaya.

My work sits at the intersection of AI security, mechanistic interpretability, and LLM infrastructure. I like building systems that make machine-learning models safer and easier to understand: red-teaming tools, security scanners, evaluation dashboards, and circuit-level interpretability of security detectors.

My current research reverse-engineers a smart-contract honeypot detector to ask a question that matters for any deployed security model: when it fires, is it reading a real signal or a quirk of the data?

Contact

• Email: rahin577@gmail.com
• GitHub: github.com/cyberjaya101
• Resume: download PDF