SALM: A Multi-Agent Framework for Language Model-Driven Social Network Simulation

TL;DR

SALM introduces a novel multi-agent framework utilizing language models for social network simulation, achieving long-term stability, efficient memory use, and validated behavioral fidelity in modeling complex social dynamics.

Contribution

The paper presents SALM, a hierarchical prompting architecture and attention-based memory system that enable stable, long-term social network simulations with reduced token usage and formal personality bounds.

Findings

Achieved stable simulation beyond 4,000 timesteps with 73% token reduction.

Attained 80% cache hit rate with sub-linear memory growth.

Validated behavioral fidelity against SNAP ego networks.

Abstract

Contemporary approaches to agent-based modeling (ABM) of social systems have traditionally emphasized rule-based behaviors, limiting their ability to capture nuanced dynamics by moving beyond predefined rules and leveraging contextual understanding from LMs of human social interaction. This paper presents SALM (Social Agent LM Framework), a novel approach for integrating language models (LMs) into social network simulation that achieves unprecedented temporal stability in multi-agent scenarios. Our primary contributions include: (1) a hierarchical prompting architecture enabling stable simulation beyond 4,000 timesteps while reducing token usage by 73%, (2) an attention-based memory system achieving 80% cache hit rates (95% CI [78%, 82%]) with sub-linear memory growth of 9.5%, and (3) formal bounds on personality stability. Through extensive validation against SNAP ego networks, we demonstrate the first LLM-based framework capable of modeling long-term social phenomena while maintaining empirically validated behavioral fidelity.