Figure 2 -.

Schematic illustrations of example computational models for social value-based decision-making and their neural correlates.

(A) Decision-making for others (Fukuda et al., 2019; Hutcherson et al., 2015). Overall value and choice probability of each option is computed by integrating information about self and other reward values. S: self; O: other; dlPFC: dorsolateral prefrontal cortex; TPJ: temporoparietal junction; and vmPFC: ventromedial prefrontal cortex.

(B) Decision-making through observing others (Burke et al., 2010; Suzuki et al., 2012). Value and choice probability of each option is computed by integrating two types of learning from others’ rewards and choices. dmPFC: dorsomedial prefrontal cortex; and IPL: inferior parietal lobule.

(C) Decision-making in strategic interactions (Hampton et al., 2008; Hill et al., 2017). Value of each option is computed by the prediction of the other’s choice that integrates learning from the other’s past choices and higher-order inference about influence of self-choice on the other’s next choice.