Figure 4.

ACC suppression significantly reduces accuracy and increases response bias. Light was delivered to ACC to suppress excitatory activity on 20% of trials through stage 4. A, Suppression reduced accuracy across stages compared with non-suppressed trials (Friedman’s, df = 1, χ² = 9.6, p = 0.002). B, Response bias increased on suppressed versus non-suppressed trials (Friedman’s, df = 1, χ² = 4.6, p = 0.033). C, Suppression did not significantly affect RT across days of training. Analyses for each stage included 4 d of training. Stage 1 included the first 4 d of training. Stage 2 included the first 4 d following performance of 65% correct responses in a 50-trial block. Stage 3 included the first 4 d following performance of 85% correct responses in a 50-trial block. Following stage 3, correction trials were reduced gradually to 0%, at which point mice received an additional 4 d of overtraining (stage 4); *p < 0.05, **p < 0.01.