Figure 9.

Belief computation in the model compared to LIP responses. The plot on the left shows responses of a model neuron representing belief (posterior probability) for leftward motion (state S_L) for stimuli moving leftward (solid) and rightward (dashed) with motion coherences 4, 8, 20, and 40% respectively. The model chose the correct action in each case. The panel on the right shows average responses of 54 neurons in cortical area LIP in a monkey (figure adapted from Roitman and Shadlen, 2002). The solid lines are for motion in the direction preferred by the neuron and dashed lines are for motion in the opposite direction. Belief responses in the model exhibit a random walk-like ramping behavior similar to LIP responses, with faster rates of growth for higher coherence values.