Padoa-Schioppa, C. | Search bioRxiv

bioRxiv2018

Neuronal Adaptation to the Value Range in Macaque Orbitofrontal Cortex

Economic choice involves computing and comparing the subjective values of different options. The magnitude of these values can vary immensely in different situations. To compensate for this variability, decision-making neural circuits adapt to the current behavioral context. In orbitofrontal cortex (OFC), neurons encode the subjective value of offered and chosen goods in a quasi-linear way. Previous work found that the gain of the encoding is lower when the value range is wider. However, previous studies did not disambiguate between neurons adapting to the value range or to the maximum value. Furthermore, they did not examine changes in baseline activity. Here we investigated how neurons in the macaque OFC adapt to changes in the value distribution. We found that neurons adapt to both the maximum and the minimum value, but only partially. Concurrently, the baseline response is higher when the minimum value is larger. Using a simulated decision circuit, we showed that higher baseline activity increases choice variability, and thus lowers the expected payoff in high value contexts.

neuroscience↗

bioRxiv2017

Neuronal adaptation and optimal coding in economic decisions

During economic decisions, neurons in orbitofrontal cortex (OFC) encode the values of offered goods. Importantly, their responses adapt to the range of values available in any given context. Prima facie, range adaptation seems to provide an efficient representation. However, uncorrected adaptation in the encoding of offer values would induce arbitrary choice biases. Thus a fundamental and open question is whether range adaptation is behaviorally advantageous. Here we present a theory of optimal coding for economic decisions. In a nutshell, the representation of offer values is optimal if it ensures maximal expected payoff. In this framework, we examine the activity of offer value cells in non-human primates. We show that their firing rates are quasi-linear functions of the offered values, even when optimal tuning functions would be highly non-linear. Most importantly, we demonstrate that for linear tuning functions range adaptation maximizes the expected payoff, even if the effects of adaptation are corrected to avoid choice biases. Thus value coding in OFC is functionally rigid (linear tuning) but parametrically plastic (range adaptation with optimal gain). Importantly, the benefit of range adaptation outweighs the cost of functional rigidity. While generally suboptimal, linear tuning may facilitate transitive choices.

neuroscience↗