Internal Model

Internal models represent a system's internal representation of its environment or a task, enabling prediction and control. Current research focuses on understanding how these models emerge in various systems, from robots and large language models to humans, employing techniques like representational similarity analysis and invertible neural networks to analyze and even influence their development. This research is significant for advancing robotics, improving human-robot interaction (particularly in rehabilitation), and enhancing our understanding of learning and cognition in both artificial and biological systems. Furthermore, efficient internal model construction is crucial for accelerating computationally intensive tasks, such as exoplanet characterization.