Ely Hahami

Ely Hahami currently serves as the Technical Director at the Harvard Computer Society, where he plays a pivotal role in steering innovative projects and fostering a vibrant community of aspiring technologists. In this leadership capacity, Ely oversees a diverse portfolio of initiatives focused on advancing research and practical applications in machine learning, deep neural networks, and mechanistic interpretability. His expertise in frameworks such as PyTorch and TensorFlow enables him to guide teams through complex model development, from designing convolutional neural networks (CNNs) for image recognition tasks to deploying natural language processing (NLP) solutions using cutting-edge techniques. Ely’s deep understanding of the theoretical underpinnings of neural networks, combined with his passion for game theory and poker, fuels his approach to tackling challenging problems in AI interpretability and robustness.

Within the Harvard Computer Society, Ely has led key projects that explore the mechanistic understanding of neural network behavior, aiming to demystify how models make decisions and improve their transparency. His proficiency with data analysis tools like Pandas and NumPy supports rigorous experimentation and visualization, while his skills in Flask and LaTeX facilitate the development of interactive dashboards and comprehensive research documentation. Ely’s interdisciplinary interests—spanning mathematics, neural network theory, and game theory—inform his approach to AI research, making him a valuable resource for students and researchers alike. As a leader, he is committed to pushing the boundaries of AI interpretability, fostering collaborative innovation, and mentoring the next generation of computer scientists dedicated to ethical and explainable AI development.