REPRESENTATION OF VALUE AND SALIENCE IN THE PRIMATE BRAINLEATHERS, MARVIN L (2015) REPRESENTATION OF VALUE AND SALIENCE IN THE PRIMATE BRAIN. Doctoral Dissertation, University of Pittsburgh. (Unpublished)
AbstractMost times, we choose goods and avoid evils. Sometimes, goods and evils choose us, which quickly grabs our attention. How do our brains represent our values so that we can choose? How do things we love or loathe automatically capture our attention? To investigate these two cognitive processes, we performed two experiments to study neuronal representations of value (worth) and salience (importance) in the primate brain. First, we tested whether neurons in LIP and amygdala encode the value or salience of choice-options early during value-based decision-making. To dissociate value from salience, we recorded neurons from LIP and the amygdala in monkeys making value-based decisions among options promising different reward-sizes and options threatening different penalty-sizes. Value increases with promised-reward but decreases with threatened-penalty. Salience increases with the intensity of both promises and threats. LIP neurons fired more for options promising more reward and also fired more for options threatening more penalty. Amygdala neurons fired more for options promising more reward but fired less for options threatening more penalty. Whereas LIP encoded salience early during decision-making, reflecting automatic capture of attention by the more important options, the amygdala encoded value early during the decision-making process. Second, we tested whether early reward-effects in neurons in LIP and amygdala depended on early automatic capture of visual attention by a more rewarding option. We dissociated cue-reward from cue-salience in LIP and amygdala using choice-options that could not acquire salience. The near-identical physical appearance of these precluded early automatic capture of attention by the large-reward option but not the monkeys’ ability to make optimal value-based decisions. LIP did not encode reward early during decision-making, only later and weakly. The number of LIP neurons with reward-effects was no greater than expected by chance. In amygdala, reward-effects were stronger and earlier than in LIP. Significant numbers of single amygdala neurons signaled the reward-size of the choice-option. Early reward-effects depend upon cue-salience in LIP but not the amygdala. The results from both experiments are consistent with the view that, early during value-based decision-making, neurons in LIP encode cue-salience whereas neurons in amygdala encode cue-value. Share
Details
MetricsMonthly Views for the past 3 yearsPlum AnalyticsActions (login required)
|