When would state-dependent learning be beneficial to an individual, and when would it be harmful?
Explain the marginal value theorem and its key assumptions. How might incorporating (a) state-dependent learning; (b) constraints on spatial memory change those assumptions?