Accounting Social Cognitive Mechanisms by the Framework of Predictive Coding and Active Inference: A Synthetic Experimental Study using Robotics Interaction Platforms

Abstract

Our group has explored possible neuropsychological mechanisms for social cognition by using predictive coding and active inference frameworks [1]. For the purpose of gaining better understanding, we have taken so-called the synthetic robotics approach wherein a set of experiments have been conducted for robot-human as well as robot-robot interactions. Especially, we examine the underlying mechanisms accounting for spontaneous coupling and decoupling among agents as well as autonomous shifts from one social context to another. We investigate also how can novel or creative behaviors be co-developed by robots and human tutors through their developmental interactive tutoring processes. Finally, I address phenomenological aspects in social cognition from our preliminary examinations on how human can feel intention or free will of the robots or how the robots can possibly do so for the humans in the human-in-the-robot-loop experiment.