We are investigating the neural mechanisms underlying object recognition, attention, and learning. We seek to understand these phenomena at multiple levels organization and use genetic tools in the mouse to probe the detailed cellular and network mechanisms. Mice are trained to perform behavioral tasks and we map the underlying neural circuit dynamics across brain areas. We measure neural activity using both optical and electrophysiological approaches, and we use optogenetics to perform perturbation experiments in order to constrain causal models of neural operation. Transgenic mice provide cell type specificity to our mechanistic interrogations. Our experiments are guided by theory, modeling, and computational analysis. We conceptualize perception as an inferential process and are developing models of hierarchical processing that include both feedforward and feedback mechanisms. We are also developing more abstract models of behavior to explain foraging and decision-making phenomena.