Newest papers

Value-Rich, Differently Organized: Measuring Value Expression and Organization in AI-Generated Debates

Maoyuan Li, Haoyuan Li, Zhongsheng Wang, Jiamou Liu · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.28604 · cited by 0 works (OpenAlex) · open-access PDF

Spanish

La evaluación actual del debate generado por LLM se centra en gran parte en la calidad argumentativa, dejando sin examinar la estructura de valor de los argumentos generados. Introducimos un marco métrico de diagnóstico para comparar cómo se expresan y organizan los valores en referencias humanas y debates generados por IA. El marco separa la expresión de valor, el rango de valores reconocidos que invoca un debate, de la organización del valor, de cómo se distribuyen y mantienen esos valores a través de turnos. Fundado en la teoría del valor básico 10 de Schwartz, representa los debates como distribuciones de valor de varios niveles sobre unidades argumentativas, declaraciones, posiciones y debates completos, permitiendo las métricas para la cobertura de valor, alineación, difusión y dinámica transversal. A través de los sistemas de gen

Machine translation (Meta M2M100); the English original follows.

Abstract

Current evaluation of LLM-generated debate largely focuses on argumentative quality, leaving the value structure of generated arguments underexamined. We introduce a diagnostic metric framework for comparing how values are expressed and organized in human-reference and AI-generated debates. The framework separates value expression, the range of recognizable values a debate invokes, from value organization, how those values are distributed and sustained across turns. Grounded in Schwartz's Basic-10 value theory, it represents debates as multi-level value distributions over argumentative units, utterances, stances, and full debates, enabling metrics for value coverage, alignment, diffusion, and cross-turn dynamics. Across multiple generation systems, AI-generated debates are broadly value-expressive, often activating as many or more value dimensions than matched human references, but they organize these values differently. Their utterances are more locally value-diffuse, and source-blind judges distinguish value clarity from value focus. These findings show that debate evaluation should consider not only value presence, but also value organization. Code and supplementary materials are available at https://github.com/Maoyuan-li/Value-Expression-and-Organization.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Value (mathematics), Argumentative, CLARITY, Epistemology, Positive economics, Metric (unit)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange