Skip to main content
Figure 1 | Behavioral and Brain Functions

Figure 1

From: Evidence that the delay-period activity of dopamine neurons corresponds to reward uncertainty rather than backpropagating TD errors

Figure 1

The delay-period activity of dopamine neurons does not appear to depend on the reward outcome of the last presentation (trial) of the same conditioned stimulus (CS). Additional analysis was performed on the data for p = 0.5 from figure 3 of Fiorillo et al [3]. (and reproduced by Niv et al [4] in their figure 1A). Rasters and histograms were generated after segregating trials according to whether the last presentation of the same CS was or was not rewarded. Both rewarded and unrewarded trials are shown. According to the hypothesis of Niv et al, one might expect to see less delay-period activity if the last presentation of the CS was followed by no reward. However, no difference was observed (see statistics in main text). A. Activity in a single dopamine neuron. B. Average activity in a population of 28 neurons selected for the presence of delay-period activity.

Back to article page