How do people learn multisensory, or amodal, representations, and what consequences do these representations have for perceptual performance? We address this question by performing a rational analysis of the problem of learning multisensory representations. This analysis makes use of a Bayesian nonparametric model that acquires latent multisensory features that optimally explain the unisensory features arising in individual sensory modalities. The model qualitatively accounts for several important aspects of multisensory perception: (a) it integrates information from multiple sensory sources in such a (...) way that it leads to superior performances in, for example, categorization tasks; (b) its performances suggest that multisensory training leads to better learning than unisensory training, even when testing is conducted in unisensory conditions; (c) its multisensory representations are modality invariant; and (d) it predicts ‘‘missing” sensory representations in modalities when the input to those modalities is absent. Our rational analysis indicates that all of these aspects emerge as part of the optimal solution to the problem of learning to represent complex multisensory environments. (shrink)

Societies are composed of groups that interact. Symbiotic groups are those in which agents complement each other in resources that they have in excess. Symbiotic groups are useful especially when the resources in an environment are distributed unevenly, because they enable agents to trade resources easily. However, for trading to happen successfully, agents in symbiotic groups need to cooperate, i.e., they should be willing to donate resources when appropriate. Similarly, if some agents in a symbiotic group are defectors, they should (...) be identified by others and eliminated from the group for the well-being of the remaining agents. Accordingly, we first study Edmonds’ tag-based model of symbiotic groups to understand the lifespan of symbiotic groups (e.g., why some groups live shorter than others). Then, we enhance Edmonds’ model by adding the capability of reciprocal interactions to agents, thus achieving a hybrid model. We capture reciprocity in three different models and study their effects on the elimination of defectors in symbiotic groups. Our experimental results show that the groups that are built with the proposed hybrid model can eliminate more defectors and earlier than tag-based models. Further, the hybrid approach can generate symbiotic groups more effectively and efficiently. (shrink)