A Neural Network Model of Conscious and Unconscious Conflict Processing

Although neural network models have been used in simulating many cognitive processes, there is no role for consciousness. The current modeling study aimed to simulate the unconscious and conscious processing based on human data from our previous functional neuroimaging study. To assess the role of conflict in conscious and unconscious control processes, we employed a hybrid connectionist model with one module for attentional control, one for interactive activation and inhibition, and one for conflict monitoring. The model successfully simulated the behavioral data, as it took longer to respond correctly to the incongruent compared with the congruent stimuli, especially in the unconscious condition. This effect was modeled through the feedback from conflict layer to modulate the response layer in the conscious condition, and the attention layer in both conscious and unconscious conditions.

Keywords:

Neural Networks , Connectionism , Cognitive Control , Conflict , Consciousness