Not yet reviewed by Pith; the record is open.

Recent experiments revealed that a certain class of inhibitory neurons in the cerebral cortex make synapses not onto cell bodies but at distal parts of dendrites of the target neurons, mediating highly nonlinear dendritic inhibition. We propose a novel form of competitive neural network model that realizes such dendritic inhibition. Contrary to the conventional lateral inhibition in neural networks, our dendritic inhibition models don't always show winner-take-all behaviors; instead, they converge to "I don't know" states when unknown input patterns are presented. We derive reduced two-dimensional dynamics for the network, showing that a drastic shift of the fixed point from a winner-take-all state to an "I don't know" state occurs in accordance with the increase in noise added to the stored patterns. By preventing misrecognition in such a way, dendritic inhibition networks achieve fine pattern discrimination, which could be one of the basic computations by inhibitory connected recurrent neural networks in the brain.