Figure 4.

Illustration of context-specific spatial learning, based on simulation data. Reward-dependent Hebbian plasticity potentiates synapses between the currently active context population and DG cells with place fields near the reward location. Left: Movement trajectory and resulting pattern of weights between “Home” context cells and DG after discovering the reward location in a “Home” trial phase. Right: Movement trajectory and resulting pattern of weights between Away context cells and DG after discovering the reward location in a Random trial phase. As weight changes in a given trial occur at only one of the two projections originating at “Home” or “Away” context cells, we used different color maps to emphasize this contextual selectivity in learning. Line width illustrates the accumulated connection strength schematically.