LATIDIA · Investigación

Fidelidad sin anclajes: un documento de posición sobre los desajustes estructurales de las métricas de fidelidad de IA explicables para la importancia de las características

arXiv: 2610.12493v1Tipo de anuncio: nuevo Resumen: Las herramientas de Inteligencia Artificial Explicable (XAI) a menudo se han presentado como métodos para aumentar la confianza en los modelos de aprendizaje automático al hacer que los aspectos de sus algoritmos sean más

La noticia

arXiv:2610.12493v1 Announce Type: new Abstract: Explainable Artificial Intelligence (XAI) tools have often been presented as methods to increase trust in machine learning models by making aspects of their algorithms more human-understandable. However, evidence accumulated over the last decade points to an overall failure of this objective, instead promoting overreliance on incorrect predictions, which can subsequently lead to algorithmic aversion. Rather, XAI appears to function primarily as a debugging tool, useful for obtaining anecdotal insights into, for example, features learned by the model. One of the main issues underlying this limitation is that XAI methods often oversimplify the inner workings of the model, resulting in misalignment between the explanation and the