HAG: Hierarchical Demographic Tree-based Agent Generation for Topic-Adaptive Simulation

TL;DR

HAG is a hierarchical framework for generating topic-adaptive agents that aligns macro-level distributions with micro-level individual attributes, improving agent-based simulation fidelity.

Contribution

The paper introduces HAG, a novel hierarchical agent generation method that combines macro-level distribution modeling with micro-level consistency, addressing limitations of existing approaches.

Findings

HAG reduces population alignment errors by 37.7%.

HAG improves sociological consistency by 18.8%.

The framework outperforms baseline methods across multiple domains.

Abstract

High-fidelity agent initialization is crucial for credible Agent-Based Modeling across diverse domains. A robust framework should be Topic-Adaptive, capturing macro-level joint distributions while ensuring micro-level individual rationality. Existing approaches fall into two categories: static data-based retrieval methods that fail to adapt to unseen topics absent from the data, and LLM-based generation methods that lack macro-level distribution awareness, resulting in inconsistencies between micro-level persona attributes and reality. To address these problems, we propose HAG, a Hierarchical Agent Generation framework that formalizes population generation as a two-stage decision process. Firstly, utilizing a World Knowledge Model to infer hierarchical conditional probabilities to construct the Topic-Adaptive Tree, achieving macro-level distribution alignment. Then, grounded real-world data, instantiation and agentic augmentation are carried out to ensure micro-level consistency. Given the lack of specialized evaluation, we establish a multi-domain benchmark and a comprehensive PACE evaluation framework. Extensive experiments show that HAG significantly outperforms representative baselines, reducing population alignment errors by an average of 37.7% and enhancing sociological consistency by 18.8%.