Satoshi Nishida

Cognitive Computational Neuroscience, Brain Decoding, Artificial Intelligence, Consciousness
Main Lab Location:
CiNet (Main bldg.)
Other Affiliations:
Guest Associate Professor, Graduate School of Frontier Biosciences, Osaka University
Mailing Address:
1-4 Yamadaoka, Suita, Osaka 565-0871, Japan

Our research has two goals. The first is to understand the neural representation of cognitive information, such as semantic and affective information, which plays an important role in determining humanity and personality. The second is to use this knowledge to develop human-like artificial intelligence (AI).

The world around us contains a large variety of complex sensory information, and each individual’s brain is unique in how it perceives the semantic and affective aspects of such information. For example, such information provides interpretations, impressions, and preferences that vary among individuals. Consequently, every person has a unique experience of the world. Our group aims to uncover the neural mechanisms underlying individual differences in semantic and affective perception by visualizing the neural representations of semantic and affective information in individual human brains. We will mathematically model the neural representations from fMRI responses to complex sensory inputs. Using this mathematical modeling, we also aim to develop brain decoding methods that reads individuals’ semantic and affective perception from fMRI responses.

Understanding the neural mechanisms underlying humanity and personality also enables us to develop human-like AI. Recent advances in AI technology are remarkable. However, the existing AI technology is not yet sophisticated enough to understand and recreate human feelings and behavior. Developing human-like AI is an important step in creating a future society in which people and AI live in harmony. For this purpose, our group aims to develop human-like AI by integrating the modeled neural representations of individual brains into the existing AI methods. We also aim at building techniques for evaluating the human-likeness of AI based on the brain decoding of human semantic and affective perception.

This research can contribute to realizing an enriched-information society that respects individual differences in personality. To facilitate this, our group actively collaborates with commercial companies for the social implementation of our research results.

Selected Publications:

Matsumoto Y†, Nishida S†, Hayashi R†, Son S, Murakami A, Yoshikawa N, Ito H, Oishi N, Masuda N, Murai T, Friston K, Nishimoto S, Takahashi H.
Disorganization of semantic brain networks in schizophrenia revealed by fMRI.
Schizophrenia Bulletin, 49(2):498–506, 2023. †Co-first

Shinkuma R, Nishida S, Maeda N, Kado M, Nishimoto S.
Reduction of information collection cost for inferring brain model relations from profile.
IEEE Transactions on Systems, Man and Cybernetics: Systems, 52(7):4057–4068, 2022.

Nishida S, Blanc A, Maeda N, Kado M, Nishimoto S. Behavioral correlates of cortical semantic representations modeled by word vectors.
PLOS Computational Biology, 17(6): e1009138, 2021.

Niikawa T, Miyahara K, Hamada HT, Nishida S.
A new experimental phenomenological method to explore the subjective features of psychological phenomena: its application to binocular rivalry.
Neuroscience of Consciousness 2020(1):niaa018, 2020.

Nishida S, Nakano Y, Blanc, A, Maeda N, Kado M, Nishimoto S.
Brain-mediated Transfer Learning of Convolutional Neural Networks.
Proceedings of the Thirty-Fourth AAAI Conference on Artificial Intelligence 34(4):5281–5288, 2020.

Nishida S, Nishimoto S.
Decoding naturalistic experiences from human brain activity via distributed representations of words
NeuroImage 180(A):232– 242, 2018.


Lab Members:

・Antoine Blanc
・Amane Tamai

・Jiaxin Wang
・Kiichi Kawahata
・Shin Okada
・Takeru Abe
・Chiyu Maeda
・Siyuan Xiang
・Haruki Takeshima