The ability to use contextual cues for the execution of appropriate behavioral responses is critical for allowing us to interact with the social environment. Recent studies suggested that a key role in this process could be played by the interaction between the ventrolateral prefrontal cortex (VLPFC) and parieto-frontal regions, such as ventral premotor area F5, to which this prefrontal sector is connected (Borra et al., 2017). In particular, it has been observed that VLPFC plays a crucial role both in action execution and action observation, and encodes task instructions in a “pragmatic” format, in terms of the goals to be achieved (Simone et al., 2017, Rozzi et al., 2023). Little is known, however, about the speci c role of prefrontal and premotor neurons in exploiting social context information to select and guide appropriate behaviors. In the current study, our aim was to assess the potential involvement of prefrontal areas 46v and 12r and of ventral premotor area F5 in the organization of the action based on social and non-social information. To address this issue, we used population approaches such as demixed principal component analysis (dPCA) and decoding analysis to investigate neural activity simultaneously recorded from these cortical sectors, in the monkey, during a task in which di erent responses were required based on the current task rule (“Imitation” or “Observation” condition), instructed by di erent colored cues, and on the presented stimulus. Neural responses were recorded under two conditions: Observation and Imitation, instructed by ambient light of two different colours, red and green, respectively. - Observation condition: The observation by the monkey of the fixation point determines the switching on of a red ambient light. After a randomized time interval, a video is presented, depicting either one of two hand actions (grasping or reaching) or one of two geometric shapes (triangle or square), which move vertically on the screen. After a further randomized time, the red light and the fixation point switch off (Go signal), instructing the monkey to briefly release the hand from the starting point. - Imitation condition: This condition is instructed by a green ambient light. The phases up to the Go signal are the same as in the Observation condition, but in this case after the Go signal the monkey had to either reach and grasp/pull or touch an object placed in front of her, based on the instruction associated to the presented video. Task-related units were identified by means of a sliding window permutation test, performed in steps of 20 ms (Model: Mixed ANOVA with two factors - Time and Type of Trial, p
