JSJae-Young Son

Jae-Young Son

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Jae-Young Son has a total of 16 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Humanlike representation of social networks in large language models

    by Jae-Young Son, Daniel Carstensen, Serra Elizabeth Favila

    Although large language models (LLMs) have made great progress in their general reasoning capabilities, reasoning about social relationships has remained a challenge. In contrast, humans are naturally skilled at reasoning about how people are connected to one another, even in large, complex social networks. Recent research in cognitive science suggests that a key component of human social intelligence is the ability to represent social networks using multistep abstraction, which integrates knowledge of both direct and indirect social connections, and therefore supports inferences about network members’ social proximity to one another. Here, we test whether LLMs represent social networks using similar computational principles. In Study 1, we converted an existing human social navigation task into text and recorded LLM responses to the same problems humans were tasked with solving. LLMs’ behaviors closely mirrored those of human participants, and were best-explained by Katz communicability, a model of multistep abstraction that captures how humans represent and navigate social networks. In Study 2, we tested whether these multistep representations persisted in a task where they offered no benefit—and were in fact detrimental to task accuracy—and found that LLMs continued to show a multistep signature even in this context. These findings suggest that multistep abstraction may be a core property of how LLMs represent relational information, closely paralleling accounts of human cognitive maps.

  • Open Access

    Toward a unified theory of social relationships and social networks

    by Jae-Young Son

    Navigating the social world requires representing how people relate to one another and how people are connected in a network. Though both are forms of relational knowledge, it is difficult to unify them in a single theoretical framework. I hypothesize that the statistics of observation may be sufficient to learn both network structure and an intuitive theory of relationships.

  • Toward a unified theory of social relationships and social networks

    by Jae-Young Son

    Navigating the social world requires representing how people relate to one another and how people are connected in a network. Though both are forms of relational knowledge, it is difficult to unify them in a single theoretical framework. I hypothesize that the statistics of observation may be sufficient to learn both network structure and an intuitive theory of relationships.

  • Open Access

    Medial temporal lobe encodes cognitive maps of real-world social networks

    by Yi Yang Teoh, Jae-Young Son, Alice Xia, Apoorva Bhandari, Oriel FeldmanHall

    Humans routinely solve social problems by navigating densely interconnected networks—gossiping strategically, brokering across cliques, and coordinating with allies. Yet, the neural representations supporting such complex navigation remain undocumented, leaving a crucial gap in our understanding of the mechanisms underlying flexible social cognition. We combine computational modeling with fMRI and a longitudinal measurement of a large, real-world social network (N = 187) to show that the medial temporal lobe encodes an abstract cognitive map of long-range connectivity in the broader network. These neural maps specifically encode the Katz communicability between network members, not merely direct ties, or Euclidean or graph distance. We additionally find that the availability of these maps in the right entorhinal cortex (rEC) and right anterior hippocampus (raHC) supports accurate inferences about information diffusion through the network. Finally, we link neural representation to consequential social change, finding that stronger rEC encoding predicts real-world brokerage that increases community cohesion over time. Together, these findings advance a general mechanism for how the brain supports flexible behavior in complex social environments, closing a major gap between theories of cognitive maps and the demands of social navigation.


  • The emotional cost of political engagement.

    by Alexander C. Walker, Jae-Young Son, Yi Yang Teoh, Oriel FeldmanHall

    Political polarization is increasingly recognized as a critical threat to individual and collective well-being. Prevailing frameworks suggest that political engagement diminishes well-being by evoking negative emotions, which act as chronic stressors. However, the relationship between politics and emotion has largely been investigated by relying on static snapshots of emotional reactions to political events, overlooking how well-being is impacted by the temporal dynamics of political engagement and associated emotional responses. Across two longitudinal experience-sampling studies that include long-form "diary" responses (N = 259, 1,788 observations), we examine how political engagement shapes daily affective experiences. Contrary to the prevailing notion that engaging with politics leads to sustained negative moods, we find that political engagement is characterized by heightened affective instability-that is, frequent and large fluctuations in affective states-which, in turn, predicts lower well-being (i.e., greater anxiety). Politically polarizing events are particularly destabilizing when they are highly salient and when individuals spontaneously engage with these events. Strong partisans on both ends of the political spectrum also show the greatest fluctuations in daily affect, characteristic of an unstable emotional life. By observing that political engagement is intimately tied to increased affective instability, this research reveals an overlooked emotional cost of political involvement. These findings open new avenues for understanding and mitigating the emotional and mental health consequences of political engagement in an era of deepening divides. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

  • Open Access

    Early insight into social network structure predicts climbing the social ladder

    by Isabella C. Aslarus, Jae-Young Son, Alice Xia, Oriel FeldmanHall

    While occupying an influential position within one’s social network brings many advantages, it is unknown how certain individuals rise in social prominence. Leveraging a longitudinal dataset that tracks an entirely new network of college freshmen ( N = 187), we test whether “climbing the social ladder” depends on knowing how other people are connected to each other. Those who ultimately come to occupy the most influential positions exhibit early and accurate representations of their network’s general, abstract structure (i.e., who belongs to which communities and cliques). In contrast, detailed, granular representations of specific friendships do not translate into gains in social influence over time. Only once the network stabilizes do the most influential individuals exhibit the most accurate representations of specific friendships. These findings reveal that those who climb the social ladder first detect their emerging network’s general structure and then fine-tune their knowledge about individual relationships between their peers as network dynamics settle.


  • Abstract Cognitive Maps for Complex Social Systems

    by Oriel FeldmanHall, Jae-Young Son, Apoorva Bhandari

    How do people represent social networks in their minds? Inspired by work on spatial navigation, recent research reveals how people use domain-general computational principles to build cognitive maps for navigating their social environments. However, some aspects of our social worlds, such as the densely interconnected networks we are embedded in—and the dynamics of information flow within them—challenge the particular construct of a Euclidean cognitive map that has evolved in the study of spatial navigation. Recent research reveals different types of abstract representations people can use to build efficient cognitive maps for navigating social networks. We argue that to solve challenges inherent to navigating social relationships (e.g., figuring out whom to trust or gossip with, building coalitions made up of weak ties), people build cognitive maps of both the direct and indirect relational ties surrounding them. Although the incorporation of indirect ties makes these maps nonveridical, their addition aids in flexible, adaptive behavior, which can be used for successfully navigating any complex social environment.


  • Open Access

    Early insight into social network structure predicts climbing the social ladder

    by Isabella Aslarus, Jae-Young Son, Alice Xia, Oriel FeldmanHall

    While occupying an influential position within one’s social network brings many advantages, it is unknown how certain individuals rise in social prominence. Leveraging a longitudinal dataset that tracks an entirely new network of college freshmen (N=187), we test whether ‘climbing the social ladder’ depends on knowing how other people are connected to each other. Those who ultimately come to occupy the most influential positions exhibit early and accurate representations of their network’s general, abstract structure (i.e., who belongs to which communities and cliques). In contrast, detailed, granular representations of specific friendships do not translate into gains in social influence over time. Only once the network stabilizes, do the most influential individuals exhibit the most accurate representations of specific friendships. These findings reveal that those who climb the social ladder first detect their emerging network’s general structure, then fine-tune their knowledge about individual relationships between their peers as network dynamics settle.

  • Open Access

    Medial temporal lobe encodes cognitive maps of real-world social networks

    by Yi Yang Teoh, Jae-Young Son, Alice Xia, Apoorva Bhandari, Oriel FeldmanHall

    People routinely navigate their complex social networks 1 : From gossiping strategically with others 2–4 to brokering connections between siloed groups 5,6 , our ability to make adaptive social choices hinges on whether we can construct useful mental representations of the social ties within our communities 7 . While decades of neuroscience research have shown that the medial temporal lobe encodes cognitive maps of physical 8–10 or conceptual space 11 , how the brain represents our social networks in the wild to solve social problems remains unknown. By combining computational models with functional neuroimaging and longitudinal measurement of an evolving and densely interconnected real-world human network (N=187), we show that the entorhinal cortex encodes a cognitive map of the long-range connectivity between pairs of network members. This social map reflects the particular demands of social navigation and is specifically formatted to encode the simultaneous connectivity between network members, which critically enables tracking how information diffuses across the network. Moreover, the strength of its encoding in the entorhinal cortex aids in brokering connections that improve cohesion within people’s social communities. Our results illuminate how a domain-general neural mechanism 12,13 is tailored to prioritize the natural dynamics of social phenomena in order to support adaptive navigation through these highly complex environments.


  • Replay shapes abstract cognitive maps for efficient social navigation

    by Jae-Young Son, Marc-Lluís Vives, Apoorva Bhandari, Oriel FeldmanHall

    To make adaptive social decisions, people must anticipate how information flows through their social network. While this requires knowledge of how people are connected, networks are too large to have first-hand experience with every possible route between individuals. How, then, are people able to accurately track information flow through social networks? Here we find that people immediately cache abstract knowledge about social network structure as they learn who is friends with whom, which enables the identification of efficient routes between remotely connected individuals. These cognitive maps of social networks, which are built while learning, are then reshaped through overnight rest. During these extended periods of rest, a replay-like mechanism helps to make these maps increasingly abstract, which privileges improvements in social navigation accuracy for the longest communication paths that span distinct communities within the network. Together, these findings provide mechanistic insight into the sophisticated mental representations humans use for social navigation.