Hidenori Tanaka

Hidenori Tanaka leads the Science of AI for Alignment group at the CBS-NTT "Physics of Intelligence" Program at Harvard University. His research integrates concepts and methods from physics, neuroscience, and psychology to advance our understanding of AI models for alignment and safety. His recent work focuses on emergent abilities in language models, multi-modal systems, and the algorithmic understanding of in-context learning. Hidenori earned his PhD in theoretical physics at Harvard University under the supervision of David Nelson and Michael Brenner, and subsequently completed a Masason Fellowship at Stanford University, working with Surya Ganguli and Daniel Fisher.