CZCristina Zuheros

Cristina Zuheros

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Cristina Zuheros has a total of 32 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    GEM: Routing-Level Behavioral Steering in Mixture-of-Experts Language Models. Sycophancy as a Case Study

    by Virilo Tejedor, Cristina Zuheros, Carlos Peláez-González, David Herrera-Poyatos, Andrés Herrera-Poyatos, Francisco Herrera

    Mixture-of-Experts (MoE) Large Language Models (LLMs) rely on sparse routing mechanisms that dynamically select specialized experts during inference. While these mechanisms are designed to improve computational efficiency, they also determine which latent competencies are expressed for a given input. We investigate whether sparse MoE routing can serve as an inference-time interface for behavioral steering, and we introduce Gating Expert Modulation (GEM) as a concrete routing intervention with which to test this hypothesis. GEM perturbs the gating logits directly, enforcing the inclusion of selected experts within the routed set and modulating their contribution, without retraining or modifying any model parameter. We use sycophancy—the tendency of a model to abandon its own stance to echo the one expressed by the user—as a case study, and instantiate GEM on Qwen3-30B-A3B. Modulating only three experts out of 6,144 reduces sycophantic stance shifts from 74.84% to 69.14% (-5.70 percentage points, a 7.6% relative reduction, 𝑝 ≪ 0.001 under McNemar’s test), and the same tuning procedure applied with the opposite objective increases them, so that the behavior can be driven in both directions from the same control surface. Across an eight-benchmark capability suite and an external evaluation on Humanity’s Last Exam, the intervention produces no observable capability degradation. These results indicate that, in the model studied, routing is not only a computational mechanism but also an inspectable and manipulable interface for post-deployment behavioral steering; how far this interface generalizes across behaviors and MoE families remains an open empirical question.

  • Open Access

    Triadic Fusion of Cognitive, Functional, and Causal Dimensions for Explainable LLMs: The TAXAL Framework

    by David Herrera-Poyatos, Carlos Peláez-González, Cristina Zuheros, Virilo Tejedor, Rosana Montes, Francisco Herrera

    Large Language Models (LLMs) are increasingly being deployed in high-risk domains where opacity, bias, and instability undermine trust and accountability. Traditional explainability methods, focused on surface outputs, do not capture the reasoning pathways, planning logic, and systemic impacts of agentic LLMs. We introduce TAXAL (Triadic Alignment for eXplainability in Agentic LLMs), a triadic fusion framework that unites three complementary dimensions: cognitive (user understanding), functional (practical utility), and causal (faithful reasoning). TAXAL provides a unified, role-sensitive foundation for designing, evaluating, and deploying explanations in diverse sociotechnical settings. Our analysis synthesizes existing methods, ranging from post-hoc attribution and dialogic interfaces to explanation-aware prompting, and situates them within the TAXAL triadic fusion model. We further demonstrate its applicability through case studies in law, education, healthcare, and public services, showing how explanation strategies adapt to institutional constraints and stakeholder roles. By combining conceptual clarity with design patterns and deployment pathways, TAXAL advances explainability as a technical and sociotechnical practice, supporting trustworthy and context-sensitive LLM applications in the era of agentic AI.

  • Open Access

    FLEX: FLEXible Federated Learning Framework

    by Francisco Herrera, Daniel Jiménez-López, Alberto Argente-Garrido, Nuria Rodríguez-Barroso, Cristina Zuheros, Ignacio Aguilera-Martos, Beatriz Bello, Mario García-Márquez, M. Victoria Luzón

    In the realm of Artificial Intelligence (AI), the need for privacy and security in data processing has become paramount. As AI applications continue to expand, the collection and handling of sensitive data raise concerns about individual privacy protection. Federated Learning (FL) emerges as a promising solution to address these challenges by enabling decentralized model training on local devices, thus preserving data privacy. This paper introduces FLEX: a FLEXible Federated Learning Framework designed to provide maximum flexibility in FL research experiments. By offering customizable features for data distribution, privacy parameters, and communication strategies, FLEX empowers researchers to innovate and develop novel FL techniques. The framework also includes libraries for specific FL implementations including: (1) anomalies, (2) blockchain, (3) adversarial attacks and defences, (4) natural language processing and (5) decision trees, enhancing its versatility and applicability in various domains. Overall, FLEX represents a significant advancement in FL research, facilitating the development of robust and efficient FL applications.

  • Open Access

    An Interpretable Client Decision Tree Aggregation process for Federated Learning

    by Alberto Argente-Garrido, Cristina Zuheros, M. Victoria Luzón, Francisco Herrera

    Trustworthy Artificial Intelligence solutions are essential in today's data-driven applications, prioritizing principles such as robustness, safety, transparency, explainability, and privacy among others. This has led to the emergence of Federated Learning as a solution for privacy and distributed machine learning. While decision trees, as self-explanatory models, are ideal for collaborative model training across multiple devices in resource-constrained environments such as federated learning environments for injecting interpretability in these models. Decision tree structure makes the aggregation in a federated learning environment not trivial. They require techniques that can merge their decision paths without introducing bias or overfitting while keeping the aggregated decision trees robust and generalizable. In this paper, we propose an Interpretable Client Decision Tree Aggregation process for Federated Learning scenarios that keeps the interpretability and the precision of the base decision trees used for the aggregation. This model is based on aggregating multiple decision paths of the decision trees and can be used on different decision tree types, such as ID3 and CART. We carry out the experiments within four datasets, and the analysis shows that the tree built with the model improves the local models, and outperforms the state-of-the-art.

  • Open Access

    Design and consensus content validity of the questionnaire for b-learning education: A 2-Tuple Fuzzy Linguistic Delphi based Decision Support Tool

    by Rosana Montes, Cristina Zuheros, Jeovani M. Morales, Noe Zermeño, Jerónimo Duran, Francsico Herrera

    Classic Delphi and Fuzzy Delphi methods are used to test content validity of data collection tools such as questionnaires. Fuzzy Delphi takes the opinion issued by judges from a linguistic perspective reducing ambiguity in opinions by using fuzzy numbers. We propose an extension named 2-Tuple Fuzzy Linguistic Delphi method to deal with scenarios in which judges show different expertise degrees by using fuzzy multigranular semantics of the linguistic terms and to obtain intermediate and final results expressed by 2-tuple linguistic values. The key idea of our proposal is to validate the full questionnaire by means of the evaluation of its parts, defining the validity of each item as a Decision Making problem. Taking the opinion of experts, we measure the degree of consensus, the degree of consistency, and the linguistic score of each item, in order to detect those items that affect, positively or negatively, the quality of the instrument. Considering the real need to evaluate a b-learning educational experience with a consensual questionnaire, we present a Decision Making model for questionnaire validation that solves it. Additionally, we contribute to this consensus reaching problem by developing an online tool under GPL v3 license. The software visualizes the collective valuations for each iteration and assists to determine which parts of the questionnaire should be modified to reach a consensual solution.

  • Open Access

    Design and consensus content validity of the questionnaire for b-learning education: A 2-Tuple Fuzzy Linguistic Delphi based Decision Support Tool

    by Rosana Montes, Cristina Zuheros, Jeovani Morales, Noe Zermeño, Jeronimo Duran, Francisco Herrera


  • Crowd Decision Making: Sparse Representation Guided by Sentiment Analysis for Leveraging the Wisdom of the Crowd

    by Cristina Zuheros, Eugenio Martinez-Camara, Enrique Herrera-Viedma, Francisco Herrera

    The “wisdom of the crowd” theory states that a nonexpert crowd makes smarter decisions than a reduced set of experts. Social network platforms are a source of evaluations in the natural language of any topic, which may be considered as the evaluations of a nonexpert crowd. Decision-making (DM) models are constrained by their inability of processing large amounts of evaluations in natural language, as those ones from social networks. We claim that evaluations from social networks can enhance the quality of multiperson multicriteria DM models. Accordingly, we propose a crowd DM model guided by sentiment analysis (SA), which solves decision situations leveraging the wisdom of the crowd available in social networks. The model uses several deep-learning SA classification models through opinion triplets to incorporate all the evaluation shades in the DM model. Likewise, the likely lack of information stemmed from the consideration of a large set of users is tackled with a sparse representation of the evaluations. We annotate and release the TripR-2020Large dataset, and we use it to evaluate the model in the use case of restaurant choice. The results show that the integration of the wisdom of the crowd and the different shades of the evaluations enhances the quality of the decision.


  • Open Access

    Sentiment Analysis based Multi-person Multi-criteria Decision Making Methodology using Natural Language Processing and Deep Learning for Smarter Decision Aid. Case study of restaurant choice using TripAdvisor reviews

    by Cristina Zuheros, Eugenio Martínez-Cámara, Enrique Herrera-Viedma, Francisco Herrera

    Decision making models are constrained by taking the expert evaluations with pre-defined numerical or linguistic terms. We claim that the use of sentiment analysis will allow decision making models to consider expert evaluations in natural language. Accordingly, we propose the Sentiment Analysis based Multi-person Multi-criteria Decision Making (SA-MpMcDM) methodology for smarter decision aid, which builds the expert evaluations from their natural language reviews, and even from their numerical ratings if they are available. The SA-MpMcDM methodology incorporates an end-to-end multi-task deep learning model for aspect based sentiment analysis, named DOC-ABSADeepL model, able to identify the aspect categories mentioned in an expert review, and to distill their opinions and criteria. The individual evaluations are aggregated via the procedure named criteria weighting through the attention of the experts. We evaluate the methodology in a case study of restaurant choice using TripAdvisor reviews, hence we build, manually annotate, and release the TripR-2020 dataset of restaurant reviews. We analyze the SA-MpMcDM methodology in different scenarios using and not using natural language and numerical evaluations. The analysis shows that the combination of both sources of information results in a higher quality preference vector.