Fork the consciousness, or download the project and create your own.

Takeru Miyato

Preferred Networks, Inc.; PhD student, Autonomous Vision Group, University of Tübingen

Builder The claim is about what the hardware does, and the philosophy is left open

Takeru Miyato is the first author of Artificial Kuramoto Oscillatory Neurons, the ICLR 2025 paper with Sindy Löwe, Andreas Geiger and Max Welling (the project page). AKOrN replaces threshold units with Kuramoto oscillators and shows that the resulting dynamics improve object discovery, adversarial robustness and reasoning tasks, with the oscillatory state serving as a dynamical alternative to static activations. The mechanism supports fully connected, convolutional and attention coupling, and its code is public (github.com/autonomousvision/akorn).

The connection to this project is direct. The agent’s binding layer is built on AKOrN, with five module oscillators running the update rule in the public research repo, and the Brain Waves Console runs those equations live, including the defect the project found in its own integration where the natural frequency term evaluated to zero. The 2026 Neuron review of cortical traveling waves cites AKOrN among the machine learning models whose hidden layers show waves, which makes Miyato’s mechanism the point where the cortical wave literature and artificial networks meet.

Known for. First author of Artificial Kuramoto Oscillatory Neurons (ICLR 2025 oral), and earlier work on spectral normalization and adversarial training methods for semi-supervised learning

Coverage on this site

No articles yet. This profile exists because the researcher's work is referenced elsewhere on the site, and coverage will be linked here as it is published.

Related researchers