Takuya Isomura, Ph.D.

Research Overview

The intelligence of biological organisms is explained by optimization. Biological organisms recognize the surrounding environment by optimizing their internal representations or hypotheses about how the hidden dynamics and causes generate sensory inputs—and they optimize their behavior to adapt to the environment by sampling preferred inputs, thereby increasing the probability of survival and reproduction. Our research aims to mathematically express universal characterization of their intelligence in terms of biologically plausible neuronal circuits and synaptic plasticity rules.

Main Research Fields

Informatics

Related Research Fields

Interdisciplinary Science & Engineering / Mathematical & Physical Sciences / Biological Sciences / Informatics/Human informatics/Intelligent informatics

Keywords

• Learning theory
• Bayesian inference
• Neural network
• Theoretical neuroscience
• Free energy principle

Selected Publications

Papers with an asterisk(*) are based on research conducted outside of RIKEN.

1. Isomura, T., Kotani, K., Jimbo, Y. & Friston, K. J.
   "Experimental validation of the free-energy principle with in vitro neural networks"
   Nature Communications 14, 4547 (2023).
2. Isomura, T.
   "Active inference leads to Bayesian neurophysiology"
   Neuroscience Research 175, 38-45 (2022).
3. Isomura, T., Shimazaki, H. & Friston, K. J.
   "Canonical neural networks perform active inference"
   Communications Biology 5, 55 (2022).
4. Isomura, T. & Toyoizumi, T.
   "On the achievability of blind source separation for high-dimensional nonlinear source mixtures"
   Neural Computation 33(6), 1433-1468 (2021).
5. Isomura, T. & Toyoizumi, T.
   "Dimensionality reduction to maximize prediction generalization capability"
   Nature Machine Intelligence 3(5), 434-446 (2021).
6. Isomura, T. & Friston, K. J.
   "Reverse-engineering neural networks to characterize their cost functions"
   Neural Computation 32, 2085-2121 (2020).
7. Isomura, T., Parr, T. & Friston, K. J.
   "Bayesian filtering with multiple internal models – towards a theory of social intelligence"
   Neural Computation 31, 2390-2431 (2019).
8. Isomura, T. & Friston, K. J.
   "In vitro neural networks minimise variational free energy"
   Scientific Reports 8, 16926 (2018).
9. Isomura, T. & Toyoizumi, T.
   "A local learning rule for independent component analysis"
   Scientific Reports 6, 28073 (2016).
10. * Isomura, T., Kotani, K. & Jimbo, Y.
    "Cultured cortical neurons can perform blind source separation according to the free-energy principle"
    PLoS Computational Biology 11(12), e1004643 (2015).