George Alvarez uses behavioral, computational, and neuroscience methods to understand how the visual system uses efficient encoding strategies to optimize the allocation of its limited resources. His research on this topic falls into four broad categories: attentional selection, memory storage, fluid resource allocation, and ensemble coding. In particular he is interested in how the visual system extracts regularities (patterns) in the flow of visual information, and uses those regularities to form more efficient representations. Several of the questions he is currently focusing on concern the neural correlates of these efficient encoding capabilities, including: (1) determining whether recall of perceptually detailed visual long-term memory episodes entails reactivation of primary visual cortex during retrieval, (2) whether learning co-variance statistics (which enables the doubling of working memory storage) alters object-representation in the ventral visual cortex, and (3) determining at which stage of visual processing the representation of ensemble statistics first emerges (because these are correlations across space, it is likely beyond v1, but where?).