Pourya Farzi

About

I investigate robust and compositional approaches for AI by integrating neural language models with symbolic and retrieval-based reasoning frameworks. My work explores adversarial robustness, explainability, and contextual grounding in large language models for high-stakes applications such as scientific discovery, health question answering, and evidence-based text generation.