Reza Ghelich is a PhD student in the School of Life Sciences at Arizona State University. Reza has a background in computational and molecular biology and is currently pursuing his PhD in Evolutionary Biology in the Geiler-Samerotte lab. His research explores intriguing parallels between emergent phenomena in genotype-phenotype-fitness landscapes and the enigmatic “black-box” behavior of deep learning models. By harnessing advanced representation learning methods, particularly Variational Autoencoders and Graph Neural Networks, he analyzes high-throughput fitness data to uncover hidden, low-dimensional structures underlying evolutionary adaptation.

Driven by a passion for mechanistic interpretability, Reza integrates techniques like saliency mapping, concept activation vectors, and feature-attribution frameworks. His goal is not merely predictive accuracy but developing models capable of revealing meaningful biological insights. By transforming opaque data into interpretable knowledge, his work aims to clarify how genetic variation translates into adaptive phenotypes across diverse environmental contexts.

Outside the lab, Reza enjoys playing guitar, jamming with friends, and hiking. His leisure time often includes thoughtful conversations on philosophy, life, and classic literature, usually accompanied by a good cup of tea!