Skip to Main Content
We demonstrate that a fixed-weight neural network can be trained with Kalman filter methods to exhibit input-output behavior that depends on which of two conditioning tasks had been performed a substantial number of time steps in the past. This behavior can also be made to survive an intervening interference task.
Neural Networks, 2003. Proceedings of the International Joint Conference on (Volume:4 )
Date of Conference: 20-24 July 2003