This course introduces students to abstract models of what and how neurons compute and concrete analyses of real neurons in action. Topics include network models of sensory processing, short- and long-term memory, reinforcement learning, the Hodgkin-Huxley model of the action potential, and techniques to analyze real experimental data. The approach will draw upon recent advances in neuroscience and deep learning. This course will emphasize students' contributions and classroom interactions. Programming homework assignments and group final projects will be a significant aspect of the course, so programming experience (Python/Matlab will be used) will be assumed. Familiarity with linear algebra and differential equations at the level of Math or Applied Math 21b will be assumed.