Yuuya Sugita
RIKEN Center for Brain Science, RIKEN, University of Würzburg
Papers
7
Total Citations
450
H-Index
6
About
Yuuya Sugita is a pioneering researcher in cognitive robotics and connectionist modeling, whose work bridges the gap between sensorimotor grounding and linguistic compositionality. His key research areas include neural network models for language acquisition, self-organizing behavior schemata, and the emergence of compositional semantics in artificial systems. Sugita's most influential contribution is the development of the Recurrent Neural Network with Parametric Biases (RNNPB), which demonstrated how distributed representations of multiple behavior schemata can self-organize in a mirror system—a concept explored in his highly cited 2004 paper (216 citations). He further advanced this line of research by introducing the second-order neural network with parametric biases (sNNPB), which simultaneously learns compositional structures from sensorimotor time series. Sugita's work is notable for its holistic approach to compositional semantics, using real robot experiments to show how language and behavior interact to generate meaning. His 2005 paper on learning semantic combinatoriality (187 citations) remains a foundational reference for researchers exploring the sub-symbolic processes underlying usage-based language acquisition. Through these contributions, Sugita has provided a compelling computational framework for understanding how compositional representations can emerge from grounded, goal-directed actions.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Simultaneously emerging Braitenberg codes and compositionality20 citations · 2011
- 4
- 5
- 6
- 7