Naoto Yoshida is a Program-Specific Researcher at Kyoto University’s Graduate School of Informatics, specializing in Symbol Emergence Systems for Qualia Structures. He earned his PhD in Information Science and Technology from the University of Tokyo in 2024. His research focuses on reinforcement learning, neural computation, personal robotics, and embodied intelligence. During his time at GROOVE X, Yoshida contributed to the LOVOT personal robot project, working on behavior design and the decision-making system. He also interned at the International Research Center for Neurointelligence at the University of Tokyo. Additionally, he has held research assistant positions at Tohoku University and OIST. In 2019, he received the WBAI Incentive Award from the Whole Brain Architecture Initiative for his contributions to brain-inspired AI research.
Graduate
MSc: Neural Computation Unit, OIST (Japan), PhD: Intelligent Systems and Informatics Lab., Information Science and Technology, The University of Tokyo (Japan)
Undergraduate
Biological Flow Studies Lab., Mechanical Engineering, Tohoku University (Japan)
Representative Publications
- Yoshida, N. (2017). Homeostatic agent for general environment. Journal of Artificial General Intelligence, 8(1), 1-22.
- Yoshida, N., Yonemura, S., Emoto, M., Kawai, K., Numaguchi, N., Nakazato, H., … & Hayashi, K. (2022). Production of character animation in a home robot: A case study of lovot. International Journal of Social Robotics, 14(1), 39-54.
- Yoshida, N., Daikoku, T., Nagai, Y., & Kuniyoshi, Y. (2024). Emergence of integrated behaviors through direct optimization for homeostasis. Neural Networks, 177, 106379.
- Yoshida, N., Kanazawa, H., & Kuniyoshi, Y. (2024). Synthesising integrated robot behaviour through reinforcement learning for homeostasis. bioRxiv, 2024-06.
Python
Reinforcement Learning
Machine Learning
Computational Neuroscience
Robotics
Your ideal research study
Construction of a machine learning theory for algorithms that enable integrated behaviour emergence based on homeostasis in robots.
What drew you to consciousness
While researching integrated behaviour emergence algorithms, I became interested in role of interoception in agents, and through recent discussions of interoceptive predictive processing, I found a connection with studies related to consciousness. In addition, I became involved in discussions of consciousness research by joining the ‘Qualia Structure’ project in Japan.
Hobbies
Reading