Neural Dynamics for Embodied Cognition (IK2023)

Dynamic Field Theory (DFT) provides a mathematical framework in which the emergence of cognition from its sensorimotor grounding can be understood. The activation dynamics of neural populations are organized as strongly recurrent neural networks that stabilize neural representations. Instabilities generate state transitions from which sequences of mental and motor acts emerge.
The tutorial will introduce the core concepts of DFT, while providing hands-on exercises and projects that make use of these concepts to build models of grounded cognition. We will discuss how DFT relates to other approaches to cognition.

Literature

  • Schöner, G.: Dynamical Systems Approaches to Cognition. In: Sun, R (ed.): The Cambridge Handbook of Computational Psychology. 2nd Edition. Cambridge University Press (in press) (pre-print)

Download Cedar Here!

15.03.2023 - 09:00-10:30

Lecture slides Slides
Exercises Introduction to Cedar/Instabilities in DFT

15.03.2023 - 14:30-16:00

Lecture slides Slides
Exercises visual search (single feature)
Exercises visual search (conjunctive)
Exercises visual search (scene memory)

16.03.2023 - 11:00-12:30

Lecture slides Language grounding lecture slides
Document Example images
Configuration files Spatial language architecture

16.03.2023 - 14:30-16:00

Lecture slides Slides
Configuration files Serial order architecture