Large Language Models (LLMs) have shown promise in simulating human behavior, yet existing agents often exhibit behavioral rigidity, a flaw frequently masked by the self-referential bias of current "LLM-as-a-judge" evaluations. By evaluating against empirical ground truth, we reveal a counter-intuitive phenomenon: increasing the intensity of prompt-driven reasoning does not enhance fidelity but rather exacerbates value polarization, collapsing population diversity. To address this, we propose the Context-Value-Action (CVA) architecture, grounded in the Stimulus-Organism-Response (S-O-R) model and Schwartz's Theory of Basic Human Values. Unlike methods relying on self-verification, CVA decouples action generation from cognitive reasoning via a novel Value Verifier trained on authentic human data to explicitly model dynamic value activation. Experiments on CVABench, which comprises over 1.1 million real-world interaction traces, demonstrate that CVA significantly outperforms baselines. Our approach effectively mitigates polarization while offering superior behavioral fidelity and interpretability.
@article{arxiv.2604.05939,
title = {Context-Value-Action Architecture for Value-Driven Large Language Model Agents},
author = {TianZe Zhang and Sirui Sun and Yuhang Xie and Xin Zhang and Zhiqiang Wu and Guojie Song},
journal= {arXiv preprint arXiv:2604.05939},
year = {2026}
}
Comments
Accepted to Findings of the Association for Computational Linguistics: ACL 2026