TY - RPRT U1 - Arbeitspapier A1 - Jaeger, Herbert T1 - Controlling Recurrent Neural Networks by Conceptors N2 - The human brain is a dynamical system whose extremely complex sensordriven neural processes give rise to conceptual, logical cognition. Understanding the interplay between nonlinear neural dynamics and concept-level cognition remains a major scientific challenge. Here I propose a mechanism of neurodynamical organization, called conceptors, which unites nonlinear dynamics with basic principles of conceptual abstraction and logic. It becomes possible to learn, store, abstract, focus, morph, generalize, de-noise and recognize a large number of dynamical patterns within a single neural system; novel patterns can be added without interfering with previously acquired ones; neural noise is automatically filtered. Conceptors help explaining how conceptual-level information processing emerges naturally and robustly in neural systems, and remove a number of roadblocks in the theory and applications of recurrent neural networks. AB - The human brain is a dynamical system whose extremely complex sensordriven neural processes give rise to conceptual, logical cognition. Understanding the interplay between nonlinear neural dynamics and concept-level cognition remains a major scientific challenge. Here I propose a mechanism of neurodynamical organization, called conceptors, which unites nonlinear dynamics with basic principles of conceptual abstraction and logic. It becomes possible to learn, store, abstract, focus, morph, generalize, de-noise and recognize a large number of dynamical patterns within a single neural system; novel patterns can be added without interfering with previously acquired ones; neural noise is automatically filtered. Conceptors help explaining how conceptual-level information processing emerges naturally and robustly in neural systems, and remove a number of roadblocks in the theory and applications of recurrent neural networks. T3 - Constructor University Technical Reports - 31 Y1 - 2014 U6 - https://nbn-resolving.org/urn:nbn:de:gbv:579-opus-1006250 UN - https://nbn-resolving.org/urn:nbn:de:gbv:579-opus-1006250 AX - 1403.3369 ER -