Kazuki Irie, PhD

Irie conducts fundamental research in deep learning to advance AI. In particular, he works on general-purpose sequence-processing neural networks and learning algorithms. His projects span a variety of domains—including supervised/unsupervised learning in language, vision, and algorithmic tasks, as well as reinforcement learning in partially observable environments such as video games. His "memory-centric deep learning" research program aims to achieve sophisticated cognitive abilities in AI, such as continual learning, (meta-)metalearning, and efficient exploration, through a unified approach of developing advanced memory algorithms that better leverage past experiences to improve future behavior. He also actively explores how advanced AI algorithms can serve as novel computational hypotheses for explaining empirical findings in cognitive science, neurobiology, and beyond. He deeply appreciates collaborating with scientists across traditional disciplinary boundaries and values the WTI spirit of "To Know, Together."

computational modeling ;  deep learning ;  high-performance computing ;  neural networks ;  language models  

artificial intelligence ;  cognitive neuroscience ;  decision-making ;  long-term memory ;  machine learning ;  synapses ;  working memory  

Kazuki Irie joined Yale as an Assistant Professor of Computer Science and Wu Tsai Investigator in July 2026. He received his PhD in Computer Science from RWTH Aachen University in Germany, after completing undergraduate and Master’s studies in Applied Mathematics at École Centrale Paris and ENS Cachan in France. Before joining Yale, he was a postdoctoral researcher in the Department of Psychology at Harvard and at the Swiss AI Lab IDSIA at the University of Lugano, Switzerland, where he also served as a lecturer teaching deep learning. His abilities in English, French, German, Italian, and Japanese sum up to 1.