Social hierarchies can be inferred through observational learning of social relationships between individuals. Yet, little is known about the causal role of specific brain regions in learning hierarchies. Here, using transcranial direct current stimulation, we show a causal role of the medial prefrontal cortex (mPFC) in learning social versus non-social hierarchies. In a Training phase, participants acquired knowledge about social and non-social hierarchies by trial and error. During a Test phase, they were presented with two items from hierarchies that were never encountered together, requiring them to make transitive inferences. Anodal stimulation over mPFC impaired social compared with non-social hierarchy learning, and this modulation was influenced by the relative social rank of the members (higher or lower status). Anodal stimulation also impaired transitive inference making, but only during early blocks before learning was established. Together, these findings demonstrate a causal role of the mPFC in learning social ranks by observation.

Reward-predictive cues can affect decision-making by enhancing instrumental responses toward the same (specific transfer) or similar (general transfer) rewards. The main theories on cue-guided decision-making consider specific transfer as driven by the activation of previously learned instrumental actions induced by cues sharing the sensory-specific properties of the reward they are associated with. However, to date, such theoretical assumption has never been directly investigated at the neural level. We hypothesize that such reactivation occurs within the premotor system and could be mapped by lateralized beta (12–30 Hz) desynchronization, a widely used marker of action selection and decision-making policy. To test this hypothesis, 42 participants (22 females) performed a pavlovian-to-instrumental transfer paradigm, while electroencephalographic activity was recorded. We anticipated increased beta desynchronization during the transfer phase when cues promoting specific transfer were presented, compared with cues predicting general transfer and neutral cues. The evidence collected confirmed our hypothesis, thus providing the first neural evidence in favor of the theorized reactivation of instrumental actions and corroborating the presence of two dissociable neural pathways underpinning specific and general transfer.

The constant challenge in social interactions involves making informed decisions in the face of competitive and cooperative dilemmas. The decision-making process can be influenced by various factors present in the social context. According to the behavior-pattern-categorization framework of information acquisition, potential biases may develop at all stages of decision-making as information about social context is progressively entered and integrated. In this study, employing the Chicken Game, we investigated the influence of varying information levels within the behavior-pattern-categorization framework (i.e., competitiveness of behavior choice, uncertainty of behavior pattern, and sociality of category) on decision-making in the dilemma of competition and cooperation. Combined with reinforcement learning models, our findings from three experiments showed that participants exhibited basic complementary behavior, becoming less competitive against highly competitive opponents and vice versa. Notably, individuals exhibited varying adaptation rates to different levels of opponent competitiveness and fluctuations. Specifically, participants adapted slower to highly competitive opponents and faster to cooperative opponents. This asymmetric adaptation in social learning is related to the rate at which various levels of information are updated. The current study disentangles the different levels of information acquisition and highlights the asymmetric processing that can occur during the updating of information within each level.

In everyday life, our behavior is often guided by environmental cues that predict rewarding or aversive outcomes. The Pavlovian-to-Instrumental Transfer (PIT) paradigm provides a framework for examining how conditioned stimuli (CS) influence instrumental actions (R) associated with specific outcomes (O). Two distinct mechanisms have been identified: specific PIT, where a cue selectively invigorates the action linked to the same outcome, and general PIT, where reward-predictive cues non-selectively enhance response vigor. Theoretical accounts propose that specific PIT depends on the reactivation of learned action representations by sensory-specific cues, yet it remains unclear whether such reactivation occurs during passive cue exposure (Pavlovian phase) and how it evolves across learning. Using fMRI (N = 31), we investigated cue-evoked neural activity during two Pavlovian learning phases, before and after instrumental learning, in a four-phase PIT paradigm. Behaviorally, participants showed robust specific and general PIT effects in the transfer phase. At the neural level, general PIT-related cues engaged occipito-temporal visual and associative regions early in learning, whereas specific PIT-related cues recruited fronto-parietal, premotor, sensorimotor and striatal regions (MFG, IPS, PMC/M1/S1, caudate, putamen) after instrumental learning. These findings indicate that Pavlovian cues dynamically engage motor and sensorimotor systems following action-outcome learning, consistent with outcome-mediated retrieval of learned action representations even in the absence of overt movement. Together, the results refine theoretical models of PIT by demonstrating learning-dependent modulation of corticostriatal circuits during passive cue processing.

Social hierarchy is a fundamental principle of social organization and an important attribute of community stability and development. Yet, little is known about the causal role of specific brain regions in learning hierarchies. Here, using transcranial direct current stimulation (tDCS), we investigated the causal role of the medial prefrontal cortex (mPFC) in learning social and non-social hierarchies. In the Training phase, participants(N=128) acquired knowledge of social and non-social hierarchy in parallel, by trial and error. During the Test phase, they were presented with two items from hierarchies that were never encountered together and required to make transitive inferences. Anodal stimulation over mPFC impaired social hierarchy learning compared with non-social learning and this modulation was influenced by the relative social rank of the members (i.e. higher or lower status). Anodal stimulation also impaired transitive inference making, but only during early blocks before learning was established. Together, our results provide causal evidence of mPFC engagement in learning social ranks by observation.

Environmental cues can influence action through outcome-related information learned elsewhere, but whether features of the instrumental acquisition history associated with those outcomes remain behaviourally relevant when the outcomes are later predicted by Pavlovian cues is unknown. Across two experiments, one conducted in the laboratory (N = 54) and one online (N = 67), we separated instrumental acquisition under low- and high-effort response requirements from Pavlovian cue learning. Participants first learned that different actions produced outcomes under relatively low or high effort requirements and separately learned cues predicting those same outcomes. Pavlovian-to-instrumental transfer (PIT) was then tested under extinction, with neither outcomes nor response requirements present. In both experiments, reward-predictive cues elicited greater response vigour than a control cue. Critically, cues predicting outcomes previously obtained under high-effort response requirements elicited greater responding than cues predicting outcomes obtained under low-effort requirements. Specific PIT was robust under both acquisition conditions, with cues selectively promoting actions associated with their predicted outcomes. There was also suggestive evidence that high-effort acquisition amplified the magnitude of specific PIT, but this modulation was not consistently supported across experiments or statistical approaches. No corresponding effect emerged for initial outcome-congruent action selection. Overall, response vigour was modulated by instrumental effort history more consistently than outcome-specific action selection, indicating that features of prior effortful instrumental experience can remain behaviourally relevant when the associated outcomes are later predicted by separately learned Pavlovian cues.