Dr. Eden Saig

Eden Saig addresses societal challenges in machine learning by combining tools and ideas from game theory, statistics, and dynamical systems. His research explores the interplay between learning algorithms and human factors, particularly economic competition and strategic behavior. While typically overlooked by traditional learning algorithms, these behavioral dynamics can fundamentally undermine the performance and reliability of learning-driven systems.

During his doctoral work in the Faculty of Computer Science at Technion–Israel Institute of Technology, Dr. Saig developed methods for analyzing and steering these effects, focusing on dynamic feedback loops induced by learning algorithms and incentive alignment in delegated machine learning tasks. Among his findings, he demonstrated how the internal structure of learning algorithms shapes the long-term composition of user populations, developed game theoretic mechanisms that steer systems toward favorable outcomes, and uncovered new foundational connections between the economic theory of contract design and the theory of statistics.

At the California Institute of Technology (Caltech), Dr. Saig is a Linde Postdoctoral Scholar, an interdisciplinary position held jointly between the departments of Computing & Mathematical Sciences and Economics. In his postdoctoral studies, he is broadening his focus to emerging challenges such as market-aware AI alignment, algorithmic monoculture in generative AI, and economic incentives in model evaluation. By developing approaches based on evolutionary game theory, statistical contract design, and social choice theory, Dr. Saig aims to ensure that machine learning systems promote economic efficiency, sustained quality, and positive societal outcomes.

Prior to his doctorate, he worked as a Research Scientist at the Facebook Core Data Science group. At the Technion, he also coordinated the Workshop in Competitive Programming, and coached the institute’s international competitive programming teams.