Fuzzy-Cognitive Integration and LLM-Assisted Interpretation: Toward Traceable and Cognitively Faithful AI Explanations

Fuente: Zenodo
Guardado en:
Detalles Bibliográficos
Autores principales: Oettgen, Walther, Saulnier, Benjamin
Formato: Recurso digital
Lenguaje:inglés
Publicado: Zenodo 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866901703256702976
author Oettgen, Walther
Saulnier, Benjamin
author_facet Oettgen, Walther
Saulnier, Benjamin
contents <p>This article presents a <strong>hybrid Fuzzy–LLM framework</strong> that bridges numerical modeling and linguistic reasoning to produce traceable, cognitively faithful explanations in sensory and cognitive modeling. Building upon the previous works (<em>Explainable Hybrid Modeling</em>, <em>Robust Explainable Modeling</em>, and <em>From Explainability to Prescription</em>), it introduces a new dataset, <strong>FuzzyCog-Sensory v1</strong>, combining physico-chemical variables, hedonic responses, and linguistic annotations.<br>The study evaluates how fuzzy rules derived from sensory data can be interpreted by <strong>Large Language Models (LLMs)</strong> under controlled traceability constraints. Quantitative metrics—<strong>Cognitive Fidelity Index (CFI)</strong>, <strong>Align-Score</strong>, and <strong>Bias Rate</strong>—measure the alignment between numerical dependencies and generated linguistic explanations. Results demonstrate that <strong>Fuzzy–LLM systems</strong> can generate consistent, auditable, and human-centered scientific narratives, advancing explainable AI toward true <strong>cognitive interpretability</strong>.</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_17386634
institution Zenodo
language eng
publishDate 2025
publisher Zenodo
record_format zenodo
spellingShingle Fuzzy-Cognitive Integration and LLM-Assisted Interpretation: Toward Traceable and Cognitively Faithful AI Explanations
Oettgen, Walther
Saulnier, Benjamin
Fuzzy Logic
Cognitive Modeling
Explainable AI
Large Language Models
Human–AI Co-Interpretation
Cognitive Fidelity
<p>This article presents a <strong>hybrid Fuzzy–LLM framework</strong> that bridges numerical modeling and linguistic reasoning to produce traceable, cognitively faithful explanations in sensory and cognitive modeling. Building upon the previous works (<em>Explainable Hybrid Modeling</em>, <em>Robust Explainable Modeling</em>, and <em>From Explainability to Prescription</em>), it introduces a new dataset, <strong>FuzzyCog-Sensory v1</strong>, combining physico-chemical variables, hedonic responses, and linguistic annotations.<br>The study evaluates how fuzzy rules derived from sensory data can be interpreted by <strong>Large Language Models (LLMs)</strong> under controlled traceability constraints. Quantitative metrics—<strong>Cognitive Fidelity Index (CFI)</strong>, <strong>Align-Score</strong>, and <strong>Bias Rate</strong>—measure the alignment between numerical dependencies and generated linguistic explanations. Results demonstrate that <strong>Fuzzy–LLM systems</strong> can generate consistent, auditable, and human-centered scientific narratives, advancing explainable AI toward true <strong>cognitive interpretability</strong>.</p>
title Fuzzy-Cognitive Integration and LLM-Assisted Interpretation: Toward Traceable and Cognitively Faithful AI Explanations
topic Fuzzy Logic
Cognitive Modeling
Explainable AI
Large Language Models
Human–AI Co-Interpretation
Cognitive Fidelity
url https://doi.org/10.5281/zenodo.17386634