Areeb User

I am an AI researcher and software engineer specializing in mechanistic interpretability, LLM analysis, and scalable backend systems. I combine rigorous experimental research with production-grade engineering to deliver reproducible results and reliable deployments.

At RespAI Lab I built automated interpretability pipelines across heterogeneous LLMs, quantified inverse-scaling effects on steerability, and mapped internal decision pathways to reveal where models override facts. In parallel, I engineered containerized microservices and orchestration for fault-tolerant, low-latency distributed systems.

I contribute to open-source projects to improve robustness, accessibility, and performance — from reducing prompt-injection vulnerabilities in large-model wrappers to improving mobile app accessibility and runtime performance — and I prioritize clean, testable code and reproducible research.