Fig. 5: TW strength encoded behavioral sensitivity to the prior reward.

A Relation between licking rate and prior trial reward. The blue and red traces show the distributions of licking rates during the pre-cue epoch (500 ms to 0 ms before cue onset) for, respectively, PR and PNR trials. The dashed vertical line is the decision criterion that afforded maximal discrimination of the prior reward based on ROC analysis (“Methods”). Pale green shading shows the trials in which the monkey licked consistently with the prior reward suggesting that they retrieved this reward, and dark gray shows trials in which they licked inconsistently with the prior reward, suggesting that they neglected the prior reward. B Relation between prior reward and PGD as a function of the monkey’s behavior. Spectrograms of GLM coefficients for the effect of the prior reward for retrieved and neglected trials; same format as Fig. 4D. C Summary plot showing the relation between TW strength in the LPFC and behavior. PGD as a function of prior trial outcome (abscissa) and licking (green vs. gray traces). Data points represent the mean and SEM of averaged PGD values in the pre-cue time–frequency interval of interest over all corresponding trials. Sensitivity to behavior is seen as an interaction, whereby the slopes are higher for retrieved versus neglected trials, which is significant for the LPFC but not PPC in both monkeys. D Relation between LFP power and behavior. z-scored LFP power (mean and SEM) in the pre-cue region of interest in the same format as (C). Analyses in panels A–D are based on trials with valid LFP and licking measurements, with the following N: LPFC Mc [2294 PR Retrieved, 1189 NPR Retrieved, 168 PR Neglected, 96 NPR Neglected]; LPFC Mj [4079 PR Retrieved, 2257 NPR Retrieved, 497 PR Neglected, 263 NPR Neglected]; PPC Mc [2788 PR Retrieved, 1148 NPR Retrieved, 200 PR Neglected, 95 NPR Neglected]; PPC Mj [1910 PR Retrieved, 650 NPR Retrieved, 241 PR Neglected, 73 NPR.