Research Papers

Counterfactual explanations explain machine learning predictions by identifying minimal input changes that would alter a model's decision. Although many existing methods successfully generate prediction-changing alternatives, they often produce unrealistic or infeasible recommendations due to a lack of explicit mechanisms for incorporating domain knowledge and intervention constraints. Neuro-symbolic AI offers a promising direction by combining data-driven predictive models with symbolic reasoning capable of representing human-understandable rules and feasible actions. This paper presents PACE, a modular neuro-symbolic framework for generating feasibility-aware counterfactual explanations. The framework separates prediction and reasoning into two components: a neural predictive model for classification and a symbolic reasoning layer that enforces domain-specific constraints during counterfactual generation. By explicitly modeling feasible interventions, the framework produces explanations consistent with domain knowledge while remaining interpretable and actionable. The approach is model-agnostic and adaptable to domains requiring realistic decision support. A case study is conducted on the Adult Income dataset, combining a multilayer perceptron classifier with Answer Set Programming (ASP) rules encoding feasible modifications to education, occupation, and working hours while preserving immutable attributes. Results highlight the trade-off between counterfactual validity and plausibility and show that symbolic constraints yield explanations that better satisfy domain-specific feasibility requirements, illustrating the potential of neuro-symbolic methods for transparent, feasibility-aware counterfactual explanation in explainable AI.

Concept Bottleneck Models (CBMs) provide an intrinsically interpretable alternative to post-hoc explanations. However, existing CBMs often rely on predefined concept vocabularies or supervised annotations, lack explicit concept grounding, and summarize each concept with a single image-level score -- discarding spatial recurrence and inter-concept dependencies. We propose a Graph-based Concept Bottleneck Model (G-CBM), an intrinsically interpretable framework that performs unsupervised concept discovery via Non-negative Matrix Factorization (NMF) and represents the discovered concepts as nodes in a per-image concept-graph representation. G-CBM matches region-level features to these concept nodes -- providing concept grounding and capturing concept recurrence across the image -- and applies a \emph{tunable concept filtering threshold} $τ$ to suppress weak region-level features. A Graph Attention Network (GAT) then performs concept-level reasoning by modeling nonlinear dependencies across nodes. Across ImageNet, HAM10000, PH2, and Derm7pt, G-CBM achieves an average relative AUC improvement of 3.7\% over a ResNet-50 baseline. Concept filtering frequently improves predictive performance while inducing selective concept use, achieving peak AUC of $0.96$ on PH2 with only 2 of 10 concepts and 0.92 on HAM10000 with 3.8 of 9 concepts. On dermoscopy benchmarks, G-CBM is competitive with supervised approaches requiring external annotations. Deletion/insertion analyses with random ablation controls show that the learned concept ranking faithfully reflects model predictions.

Artificial intelligence governance is consolidating around three structurally heterogeneous instruments: ISO/IEC 42001:2023, the first certifiable Artificial Intelligence Management System (AIMS) standard; the United States NIST AI Risk Management Framework (AI RMF 1.0), a voluntary, socio-technical risk model; and the European Union Artificial Intelligence Act (Regulation (EU) 2024/1689), a binding, risk-tiered law. Although these instruments share the goal of trustworthy AI, they differ fundamentally in legal status, governance subject, and conception of risk, so that the control-level crosswalks now common in practice are both incomplete and, in places, misleading. Drawing on document analysis of the official standards and frameworks and on comparative governance literature, this article reconciles the three instruments into a Unified AI Governance Taxonomy (UAGT) organized as five analytical layers -- normative purpose, governance subject, risk logic, control architecture, and evidence and assurance -- bound by a traceability spine and expressed through eight regulation-stable governance domains. The taxonomy is deliberately current, incorporating the May 2026 Digital Omnibus amendments to the AI Act and the 2024 NIST Generative AI Profile. A four-step implementation model and two worked high-risk examples -- an AI-enabled clinical decision-support system and an AI credit-scoring system -- show how organizations can operate a single control library and evidence base that supports ISO/IEC 42001 certification, NIST AI RMF adoption, and EU AI Act compliance without duplicative effort. We are explicit about where unification breaks down: the non-fungibility of legal conformity and voluntary certification, the mismatch between list-based and contextual risk ontologies, enforcement asymmetry, and the widening gap between all three instruments and general-purpose and agentic AI.

The growing demand for privacy-preserving data sharing has positioned synthetic data generation as a critical component of responsible AI workflows. Despite notable advances in generative modeling, existing solutions often lack integration of adaptive generation strategies, multi-metric evaluation, and accessible end-to-end generators within a unified web-based toolkit. In this work, we introduce TDGT (Tabular Data Generation Toolkit), a web-based toolkit for synthetic tabular data generation and fidelity assessment. TDGT introduces the Adaptive Bayesian Mixture Synthesizer (ABMS), a novel algorithm that autonomously determines the optimal number of mixture components through iterative cluster quality optimization, eliminating the need for manual hyperparameter configuration. Building upon ABMS, we further propose VAE-ABMS, a hybrid architecture that couples Variational Autoencoder-based latent space learning with adaptive Bayesian mixture synthesis, enabling high-fidelity generation of complex, nonlinear tabular distributions. For large-scale scenarios, TDGT provides a GPU-accelerated variant of ABMS leveraging CUDA-based k-means clustering and Gaussian mixture fitting. Synthetic data fidelity is assessed through eleven statistical fidelity metrics spanning distributional divergence, structural correlation, and sample-level similarity, complemented by privacy risk indicators including k-anonymity scoring and disclosure rate estimation. The web-based toolkit supports a real-time streaming interface with interactive Plotly-based visualizations. TDGT is assessed across datasets from healthcare, socioeconomic modeling, and cybersecurity domains, demonstrating consistent generation fidelity and statistical coherence across heterogeneous feature types and data scales.

Flaw reporting for deployed AI systems is fundamental to identifying system failures and improving AI safety. Yet the AI reporting ecosystem is fragmented: researchers who identify flaws often do not know what or where to report, and groups who receive reports rarely share them with other relevant stakeholders. As a result, good-faith reporters duplicate effort by submitting many different forms, and recipients lack standardized, triage-ready information. We audit 12 reporting systems published by AI developers, cybersecurity groups, and AI flaw aggregators, identifying five recurring design challenges spanning discoverability, scope, information collection, coordination, and guidance for strict-liability cases. Building on this analysis and feedback from 49 experts across 32 organizations representing developers, security researchers, and ecosystem coordinators, we introduce FLARE-AI, an open-source AI flaw reporting system designed for interoperability with existing systems. FLARE-AI streamlines flaw report creation by collecting triage-relevant information through conditional logic and early classification, then enables optional dissemination of standardized, machine-readable reports to multiple developers, coordinators, and incident registries from a single submission. By lowering barriers to reporting AI flaws and improving interoperability across stakeholders, FLARE-AI helps break down silos and accelerate remediation across the AI ecosystem.

Scientific Explanations in Health Sciences: Causality, Trust, and Epistemic Adequacy

Martina Mattioli, Marcello Pelillo

Medical Artificial Intelligence (AI) is widely expected to transform clinical practice, yet the decision-making processes of many Machine Learning (ML) models remain opaque. Explainability has been advanced as a partial remedy to clarify why AI generates predictions, particularly in high-stakes contexts. Despite ongoing efforts, debates on what constitutes an adequate medical explanation remain unsettled. Yet, explanation has long been a central topic of inquiry in the philosophy of science and medicine. The insights developed in these fields, however, have been largely overlooked in contemporary explainable AI (XAI) research, leaving its foundational assumptions insufficiently examined. To address this gap, this paper develops a critical review at the intersection of philosophy of science and XAI. It examines prevailing accounts of what counts as an explanation in the health sciences and assesses their adequacy for informing XAI in medicine, arguing that they provide necessary conditions for a philosophically grounded approach to explainability in this domain. Building on this foundational philosophical literature, the discussion identifies three central axes of analysis: the role of causality in medical reasoning, the epistemic and relational dimensions of medical trust, and the criteria of explanatory adequacy as shaped by the pragmatic needs of diverse stakeholders. By integrating philosophical analysis with current developments in medical AI, the paper outlines principles for designing XAI systems that offer explanations that are not only epistemically robust but also aligned with the epistemic and practical requirements of clinical decision-making, shaping ongoing debates in medical XAI toward underexplored conceptual foundations.

A Fair and Transparent Framework for Speech-Based Depression Detection: Balancing Interpretability and Performance

Mariel Estevez, Alfonso Ortega, Antonio Miguel +1

While speech provides rich, non-invasive biomarkers for mental-health assessment, clinical adoption is limited by opaque models and potential demographic bias. In this work we propose a methodological framework to evaluate robustness and interpretability for automated depression detection on the extended DAIC-WOZ dataset using low-complexity machine learning baselines (RF, SVM, and MLP) chosen to mitigate overfitting and enhance generalization in combination with human-understandable acoustic features (MFCCs, eGeMAPS). To balance accuracy with clinical trust, we leverage explainability methods (LIME and SHAP) for feature selection, validating our findings with statistical significance tests and demographic fairness analyses to mitigate spurious, artifact-driven correlations. Empirical results demonstrate that an optimized subset of explainable AI (XAI)-selected features combined with an MLP architecture achieves a state-of-the-art test accuracy of 82\%. Ultimately, this work provides a transparent framework for robust and ethical assistive technologies that can be applied to any other binary task.

Hallucination detection has become a pressing requirement for trustworthy AI deployment at scale. The most accurate detection methods depend on GPU-intensive inference, proprietary API calls, or white-box access to the generating model, putting them out of reach for resource-constrained researchers and practitioners. We explore a practical alternative: how well can hallucination detection perform using only lightweight, CPU-feasible methods built on public models? We benchmark four such detectors, ROUGE-L, semantic similarity, BERTScore, and a Natural Language Inference (NLI) detector based on a FEVER-trained DeBERTa model, together with a score-level ensemble of similarity and NLI. We evaluate them across all three tasks of the HaluEval benchmark: question answering (QA), dialogue, and summarisation. We calibrate on a held-out validation split, evaluate on 2,000 test instances per task, and report bootstrap confidence intervals. The similarity-NLI ensemble is the most consistent method, but absolute performance is highly task-dependent. It ranks best on QA (F1 = 0.792, AUC-ROC = 0.873) and on dialogue (F1 = 0.694, AUC-ROC = 0.749), where NLI is the strongest standalone method; on summarisation every method performs near chance (AUC-ROC between 0.469 and 0.574). We then ask whether that failure is intrinsic to lightweight detection or an artifact of our single-pass design, and find it is largely the latter. Raising the premise budget from 800 to 1600 characters lifts summarisation AUC-ROC from 0.567 to 0.629, and replacing single-pass scoring with sentence-level chunk aggregation reaches 0.683, still on CPU with the same model, though at roughly twenty times the NLI inference. Summarisation remains by far the hardest task, but our results do not support treating lightweight detection as intrinsically unsuited to it.

Artificial intelligence is increasingly used in medical imaging, yet its robustness to input perturbations remains a critical concern for a wide clinical adoption. To this end, we used adversarial examples to systematically probe vulnerabilities of a U-Net-based model observer for computed tomography protocol optimization, performing detection and localization of low-contrast objects in a phantom dataset. Adversarial attacks were generated using both gradient-based and optimization-based white-box methods. Fast gradient perturbations produced high misclassification rates, reaching up to 75% at intermediate perturbation levels while remaining visually imperceptible. Localization was more robust, with success rates of about 25% for small perturbations and 42% at moderate levels. In contrast, optimization-based attack achieved success rates close to 50% for both tasks. To mitigate these vulnerabilities, dynamic adversarial training was implemented. This reduced the success rate of optimization-based attacks to 7% for classification and 13% when including localization-specific training, demonstrating a substantial robustness improvement without compromising task performances, confirmed by localization receiver operating characteristic analysis. To further interpret model behavior, radiomic texture analysis was performed on original and adversarial images. While most global image statistics remain stable, specific texture-related features exhibit consistent changes in successful attacks, highlighting the model's sensitivity to subtle local intensity patterns. Overall, adversarial training improves robustness without degrading performance, while radiomic analysis reveals interpretable links between texture alterations and prediction failures, supporting more reliable and explainable AI systems for medical imaging.

This paper presents a systematic survey and conceptual synthesis of the shared measurement problem underlying large language model (LLM) evaluation and AI safety: benchmark scores, reward signals, and safety metrics can improve while the capabilities and alignment properties they are meant to represent remain uncertain. Synthesizing 373 primary studies published between 2018 and 2026, the survey organizes evidence on benchmark validity, contamination, dynamic evaluation, LLM-as-a-judge protocols, adversarial safety testing, reward and proxy optimization, mechanistic interpretability, and AI governance into an eight-stream evidence taxonomy. Building on this synthesis, we introduce EvalSafetyGap, a conceptual framework that unifies benchmark-validity and alignment-failure research as a shared proxy-target divergence problem under optimization pressure, formalized through a Goodhart-inspired Instability Decomposition and an Alignment Trilemma. An exploratory ten-model public-evidence audit illustrates the framework by showing why capability, behavioral robustness, and governance disclosure should be reported as separate evidence layers rather than collapsed into a single safety score. The survey closes with a research agenda for dynamic and contamination-resistant benchmarks, pre-specified multi-attempt threat models, version-locked evaluation, transparent source reporting, and validated mechanistic safety indicators, offering researchers, model developers, and AI auditors a shared vocabulary for measurement-aware LLM safety evaluation.

From housing allocation for households experiencing homelessness to triage in emergency departments, LLMs are increasingly being considered as judges of consequential decisions that require ranking people for scarce resources. Ranking large groups simultaneously is cognitively demanding and error-prone. A natural solution, drawing on decades of social choice theory, elicits pairwise comparisons and aggregates them into a total order. However, a fundamental question remains when LLMs serve as the pairwise judge: how can a practitioner tell, before committing to a ranking, whether the LLM's judgments are sufficiently consistent to trust the result? We discuss two different ways of identifying consistency. A classical diagnostic, the coefficient of consistency $ζ$, originally developed to measure judge reliability by counting circular triads in tournament graphs, provides a cheap, model-free measure of intra-run consistency. Various standard measures of distance between rankings, for example Kendall's $τ$, can measure inter-run variability. We show, in both theory and practice, that these measures are independently valuable, and advocate for using both to assess reliability of rankings. We demonstrate the practical importance of our results across two high-stakes prioritization tasks: homelessness service allocation and emergency department triage. Three different leading LLMs have considerably different performance profiles across these two axes of consistency. We provide guidelines for how practitioners could think about measuring and assessing consistency before committing to a model for ranking or prioritization.

Speech conveys rich emotional information. As Speech Emotion Recognition (SER) is usually deployed in privacy-sensitive and reliability-critical environments, adversarial attacks on SER have attracted increasing attention. Existing sparse attacks control the number of perturbed elements, yet, they often lack explainability guidance and explicit measures of explanation consistency. A unified treatment of sparsity and magnitude constraints is also uncommon. In addition, transferability across attack families and target models remains limited. Hence, we propose a SalIency-Guided sparse Mask Attack (SIGMA). On self-supervised speech features, we use post-hoc explainable artificial intelligence (XAI) techniques to produce saliency maps and identify the scope of the mask, and then restrict magnitude-bounded updates to this mask. The mask is computed once and can be reused across models and different sparsity attacks to amortise cost. We evaluate on the IEMOCAP and TESS datasets. Under matched budgets and across multiple sparse-attack settings, SIGMA maintains competitive attack success rates, navigating a conscious trade-off between attack efficacy and explanation consistency. SIGMA therefore provides an efficient and interpretable framework for analysing the vulnerability and explanation behaviour of SER models under structured perturbations.