“Nothing in life is to be feared, it is only to be understood. Now is the time to understand more, so that we may fear less.”
Ultimately, the success of Physics of Intelligence should be measured by its ability to forecast the future. We aim to make concrete, qualitative predictions about emerging AI safety risks, drawing on a scientific foundation of intelligence to make the unimaginable intelligible.
Currently exploring
This post opens our series on Mechanistic Swarm Interpretability: understanding how the personas of individual agents and their interactions shape collective personas and behavior.
Read the blog
Yongyi Yang, Core Francisco Park, Ekdeep Singh Lubana, Maya Okawa, Wei Hu, Hidenori Tanaka
Ekdeep Singh Lubana, Kyogo Kawaguchi, Robert P. Dick, Hidenori Tanaka
Maya Okawa, Ekdeep Singh Lubana, R. P. Dick, Hidenori Tanaka
Hidenori Tanaka, A. Nayebi, N. Maheswaranathan, L. McIntosh, S. A. Baccus, Surya Ganguli
February 2026
Hidenori Tanaka joined the Cambridge Boston Alignment Initiative Summer Research Fellowship in AI Safety as a mentor.
October 2025
NTT Research announced that Dr. Mai Uchida will collaborate with the Physics of AI Group on psychiatry, mental health, and safe human-AI systems.
April 2025
Including studies on competition dynamics in in-context learning, percolation models of emergence, and concept learning dynamics.
November 2024
Hidenori Tanaka joined the ML Alignment & Theory Scholars (MATS) program as a mentor.
