From Who They Are to How They Act: Behavioral Traits in Generative Agent-Based Models of Social Media

TL;DR

This paper addresses the limitation of Generative Agent-Based Models (GABM) in social media by introducing explicit behavioral traits that regulate agents' propensities across posting, re-sharing, commenting, reacting, and inactivity. The authors extend a GABM framework with a two-layer agent profile (identity + behavioral traits), a three-part memory system (STM, LTM, AM), and a reasoning module to produce action choices, enabling propagation chains through an enhanced content propagation mechanism. Through 980 agents and multiple configurations, they show that behavioral traits sustain heterogeneous, profile-consistent participation, enable realistic diffusion via amplification- and interaction-oriented profiles, and yield network centrality patterns aligned with real-world data. The work, including empirical grounding with a 2020 Twitter dataset and public code release, significantly advances GABM as a tool for studying online information diffusion and social dynamics, with implications for understanding and moderating platform phenomena.

Abstract

Generative Agent-Based Modeling (GABM) leverages Large Language Models to create autonomous agents that simulate human behavior in social media environments, demonstrating potential for modeling information propagation, influence processes, and network phenomena. While existing frameworks characterize agents through demographic attributes, personality traits, and interests, they lack mechanisms to encode behavioral dispositions toward platform actions, causing agents to exhibit homogeneous engagement patterns rather than the differentiated participation styles observed on real platforms. In this paper, we investigate the role of behavioral traits as an explicit characterization layer to regulate agents' propensities across posting, re-sharing, commenting, reacting, and inactivity. Through large-scale simulations involving 980 agents and validation against real-world social media data, we demonstrate that behavioral traits are essential to sustain heterogeneous, profile-consistent participation patterns and enable realistic content propagation dynamics through the interplay of amplification- and interaction-oriented profiles. Our findings establish that modeling how agents act-not only who they are-is necessary for advancing GABM as a tool for studying social media phenomena.

Figures (14)

• Figure 1: Overview of the proposed GABM-based simulation framework. The framework models large-scale social media dynamics through a population of generative agents, each defined by a two-layer profile of identity traits and behavioral traits. Agents rely on a three-part memory unit (short-term, long-term, and activity memory) to autonomously select and perform platform actions (e.g., posting, following, re-sharing, reacting) until a stop condition is reached. In the figure, the main contributions of this work are highlighted: the introduction of behavioral traits and the activity memory (AM) component, as well as an extended re-sharing mechanism that allows agents to re-share already re-shared content, enabling content propagation through re-sharing chains.
• Figure 2: (a) Average action probabilities in the FullModel and IdentityOnly (behavioral traits disabled) configurations. (b) Average action probabilities across the seven behavioral traits. (c) Average action probabilities across the ten OCEAN traits. (d) Comparison of cluster distributions in the FullModel and IdentityOnly configurations. With traits enabled, agents form diverse, profile-consistent clusters, while without traits, most collapse into a single cluster.
• Figure 3: Comparison of content production and amplification dynamics for the FullModel configuration, illustrating (a) the temporal evolution of first- and second-order actions and (b) the cumulative percentage of original and re-shared content.
• Figure 4: Distribution of behavioral traits across positions in propagation chains. Position 0 corresponds to the creation of original content, while subsequent positions represent successive re-shares. Bars indicate the percentage of agents from each behavioral trait contributing at each stage.
• Figure 5: a) In-degree and b) out-degree centralities across behavioral traits in the re-sharing network.