A single dynamical property can account for the capacity to learn, from artificial networks to the mammalian brain.
Ravi Chopra, Juntao Zhong, Ezra S Miller, Gemechu Bekele Tolossa, Leandro J. Fosque, Jordan A. Janzen-Meza, Nicholas W. DeKorver, Réjean M. Guerriero, Neil J. Ritter, Mary E. Lambo, Kiran Bhaskaran‐Nair, Stephen D. Van Hooser, Woodrow L. Shew, Keith B. Hengen
The paper shows that proximity to criticality in neural systems predicts learning rates and cognitive abilities across species.
bioRxiv · Jul 14, 2026 · PreprintNeuroscience