Figure 4. Cocaine exposure reduced delay representation, and activity was discounted for long-delay rewards in reward-responsive cells. A, Single neuron example during a long-delay trial in which activity is sustained during the delay period. Activity is aligned to well entry (time 0, black line). B, Single neuron example during a long-delay trial in which activity initially increases after well entry and subsequently decreases, failing to maintain reward representation throughout the long delay. C, D, Percentage of reward-selective cells from control (C) and cocaine-exposed (D) rats, as defined by significantly greater firing during the reward analysis epoch (250 ms before reward to 1 s after reward delivery) compared with baseline (1 s before odor onset; Wilcoxon, p < 0.05). E, F, Normalized firing rates for neurons that increased firing rates during the reward epoch from control (n = 144; E) and cocaine-exposed rats (n = 265; F). Activity is aligned to fluid well entry, comparing short (blue dotted line) and long-delay (red line) rewards. G, H, Normalized firing rates for cells from control (G) and cocaine-exposed (H) rats, aligning activity for short (blue), long (red), big (green), and small (orange) rewards to reward delivery. I, J, Distribution of value indices for cells taken from control (I) and cocaine-exposed (J) rats, comparing firing rates during the first and last 500 ms of long delay trials. K, L, Distribution of delay indices (short – long / short + long) for cells taken from control (K) and cocaine-exposed (L) rats, comparing firing rates during the reward epoch. M, N, Distribution of context indices reflecting changes in global value (size block – delay block / size block + delay block) comparing average activity during size and delay blocks during the postreward delivery epoch (1–2 s after reward delivery), for control (M) and cocaine-exposed (N) rats. Black bars represent counts of neurons with firing that significantly differed between trial types (p < 0.05, Wilcoxon).