Proceedings of the Annual Meeting of the Cognitive Science Society

Despite the tremendous advances of Artificial Intelligence, a general theory of intelligent systems, connecting the psycho-logical, neuroscientific and computational levels is lacking. Artificial Neural Networks are good starting points to buildthe theory. We propose to analyze generalization of learning in simple but challenging problems. We have previouslyproposed to concentrate on learning sameness, as we have shown that this is difficult for a SRN. Here we present theresults of trying to use a Long-Short Term Memory Network to learn sameness. We show that the LSTM although muchmore efficient to learn partial examples of sameness fails to generalize to a proportion of the examples. This suggests thatLSTM and SRN share a core set of features that make generalization of sameness problematic. By analyzing where thetwo models fail, we arrive at a proposal of what makes sameness hard to learn and generalize in recurrent neural networks.