arXiv subjects

Eirini Ntoutsi

Publications and source records attributed to Eirini Ntoutsi.

At least 19 recordsLinked to original sources

arXiv2026

On Adversarial Vulnerability of Vision-Language Models through the Lens of Intermediate Spectral Subspaces

Adversarial vulnerability in deep neural networks (DNNs) has been studied from the perspectives of decision-boundary geometry, feature robustness, input-output Jacobians, and the instability of inverse problems. Here, we focus on the spectral structure of intermediate linear transformations that propagate information through modern DNNs, an unexplored mechanism of adversarial vulnerability. Specifically, we investigate transformer-based vision-language models, whose linear layers admit interpretable spectral decompositions and whose widespread adoption makes understanding their robustness increasingly important. We propose a white-box spectral-subspace-guided attack (SSGRA) that aligns intermediate representations with the subspace spanned by the bottom right singular vectors. Our experiments show improved attack effectiveness over existing baselines. In addition, SSGRA offers a spectral interpretation of adversarial vulnerability in VLMs, providing insights for improving their robustness.

cs.LG↗

arXiv2026

Revealing Hidden Vulnerabilities in Autoencoders through Gradient Signal Restoration

Adversarial robustness of deep autoencoders (AEs) has received less attention than that of discriminative models, although their compressed latent representations induce ill-conditioned mappings that can amplify small input perturbations and destabilize reconstructions. Existing white-box attacks for AEs, which optimize norm-bounded adversarial perturbations to maximize reconstruction damage, often converge to suboptimal perturbations, thereby potentially overstating AE robustness. We show that this limitation is linked to vanishing adversarial loss gradients during backpropagation through ill-conditioned layers, associated with near-zero singular values in their intermediate weight matrices. To address this, we propose GRILL (Gradient Signal Restoration in Ill-Conditioned Layers), a framework designed to mitigate gradient degradation and improve the reliability of adversarial robustness evaluation in encoder-decoder architectures. GRILL is designed to mitigate adversarial gradient degradation during optimization, enabling attacks to better approximate high-distortion perturbations under fixed norm constraints. Through extensive experiments across multiple AE architectures, under both sample-specific and universal attacks, as well as standard and adaptive attack settings, we show that GRILL significantly increases attack effectiveness, thereby exposing vulnerabilities hidden by existing attack limitations. Beyond AEs, we provide preliminary evidence that modern multimodal encoder-decoder architectures exhibit similar vulnerabilities.

cs.LG↗

arXiv2025

Mitigating Semantic Drift: Evaluating LLMs' Efficacy in Psychotherapy through MI Dialogue Summarization

Recent advancements in large language models (LLMs) have shown their potential across both general and domain-specific tasks. However, there is a growing concern regarding their lack of sensitivity, factual incorrectness in responses, inconsistent expressions of empathy, bias, hallucinations, and overall inability to capture the depth and complexity of human understanding, especially in low-resource and sensitive domains such as psychology. To address these challenges, our study employs a mixed-methods approach to evaluate the efficacy of LLMs in psychotherapy. We use LLMs to generate precise summaries of motivational interviewing (MI) dialogues and design a two-stage annotation scheme based on key components of the Motivational Interviewing Treatment Integrity (MITI) framework, namely evocation, collaboration, autonomy, direction, empathy, and a non-judgmental attitude. Using expert-annotated MI dialogues as ground truth, we formulate multi-class classification tasks to assess model performance under progressive prompting techniques, incorporating one-shot and few-shot prompting. Our results offer insights into LLMs' capacity for understanding complex psychological constructs and highlight best practices to mitigate ``semantic drift" in therapeutic settings. Our work contributes not only to the MI community by providing a high-quality annotated dataset to address data scarcity in low-resource domains but also critical insights for using LLMs for precise contextual interpretation in complex behavioral therapy.

cs.CL↗

arXiv2025

Effector: A Python package for regional explanations

Effector is a Python package for interpreting machine learning (ML) models that are trained on tabular data through global and regional feature effects. Global effects, like Partial Dependence Plot (PDP) and Accumulated Local Effects (ALE), are widely used for explaining tabular ML models due to their simplicity -- each feature's average influence on the prediction is summarized by a single 1D plot. However, when features are interacting, global effects can be misleading. Regional effects address this by partitioning the input space into disjoint subregions with minimal interactions within each and computing a separate regional effect per subspace. Regional effects are then visualized by a set of 1D plots per feature. Effector provides efficient implementations of state-of-the-art global and regional feature effects methods under a unified API. The package integrates seamlessly with major ML libraries like scikit-learn and PyTorch. It is designed to be modular and extensible, and comes with comprehensive documentation and tutorials. Effector is an open-source project publicly available on Github at https://github.com/givasile/effector.

cs.LG↗

arXiv2025

A Deep Latent Factor Graph Clustering with Fairness-Utility Trade-off Perspective

Fair graph clustering seeks partitions that respect network structure while maintaining proportional representation across sensitive groups, with applications spanning community detection, team formation, resource allocation, and social network analysis. Many existing approaches enforce rigid constraints or rely on multi-stage pipelines (e.g., spectral embedding followed by $k$-means), limiting trade-off control, interpretability, and scalability. We introduce \emph{DFNMF}, an end-to-end deep nonnegative tri-factorization tailored to graphs that directly optimizes cluster assignments with a soft statistical-parity regularizer. A single parameter $λ$ tunes the fairness--utility balance, while nonnegativity yields parts-based factors and transparent soft memberships. The optimization uses sparse-friendly alternating updates and scales near-linearly with the number of edges. Across synthetic and real networks, DFNMF achieves substantially higher group balance at comparable modularity, often dominating state-of-the-art baselines on the Pareto front. The code is available at https://github.com/SiamakGhodsi/DFNMF.git.

cs.LG↗

arXiv2025

TABFAIRGDT: A Fast Fair Tabular Data Generator using Autoregressive Decision Trees

Ensuring fairness in machine learning remains a significant challenge, as models often inherit biases from their training data. Generative models have recently emerged as a promising approach to mitigate bias at the data level while preserving utility. However, many rely on deep architectures, despite evidence that simpler models can be highly effective for tabular data. In this work, we introduce TABFAIRGDT, a novel method for generating fair synthetic tabular data using autoregressive decision trees. To enforce fairness, we propose a soft leaf resampling technique that adjusts decision tree outputs to reduce bias while preserving predictive performance. Our approach is non-parametric, effectively capturing complex relationships between mixed feature types, without relying on assumptions about the underlying data distributions. We evaluate TABFAIRGDT on benchmark fairness datasets and demonstrate that it outperforms state-of-the-art (SOTA) deep generative models, achieving better fairness-utility trade-off for downstream tasks, as well as higher synthetic data quality. Moreover, our method is lightweight, highly efficient, and CPU-compatible, requiring no data pre-processing. Remarkably, TABFAIRGDT achieves a 72% average speedup over the fastest SOTA baseline across various dataset sizes, and can generate fair synthetic data for medium-sized datasets (10 features, 10K samples) in just one second on a standard CPU, making it an ideal solution for real-world fairness-sensitive applications.

cs.LG↗

arXiv2025

MMM-fair: An Interactive Toolkit for Exploring and Operationalizing Multi-Fairness Trade-offs

Fairness-aware classification requires balancing performance and fairness, often intensified by intersectional biases. Conflicting fairness definitions further complicate the task, making it difficult to identify universally fair solutions. Despite growing regulatory and societal demands for equitable AI, popular toolkits offer limited support for exploring multi-dimensional fairness and related trade-offs. To address this, we present mmm-fair, an open-source toolkit leveraging boosting-based ensemble approaches that dynamically optimizes model weights to jointly minimize classification errors and diverse fairness violations, enabling flexible multi-objective optimization. The system empowers users to deploy models that align with their context-specific needs while reliably uncovering intersectional biases often missed by state-of-the-art methods. In a nutshell, mmm-fair uniquely combines in-depth multi-attribute fairness, multi-objective optimization, a no-code, chat-based interface, LLM-powered explanations, interactive Pareto exploration for model selection, custom fairness constraint definition, and deployment-ready models in a single open-source toolkit, a combination rarely found in existing fairness tools. Demo walkthrough available at: https://youtu.be/_rcpjlXFqkw.

cs.LG↗

arXiv2025

Achieving Hilbert-Schmidt Independence Under Rényi Differential Privacy for Fair and Private Data Generation

As privacy regulations such as the GDPR and HIPAA and responsibility frameworks for artificial intelligence such as the AI Act gain traction, the ethical and responsible use of real-world data faces increasing constraints. Synthetic data generation has emerged as a promising solution to risk-aware data sharing and model development, particularly for tabular datasets that are foundational to sensitive domains such as healthcare. To address both privacy and fairness concerns in this setting, we propose FLIP (Fair Latent Intervention under Privacy guarantees), a transformer-based variational autoencoder augmented with latent diffusion to generate heterogeneous tabular data. Unlike the typical setup in fairness-aware data generation, we assume a task-agnostic setup, not reliant on a fixed, defined downstream task, thus offering broader applicability. To ensure privacy, FLIP employs Rényi differential privacy (RDP) constraints during training and addresses fairness in the input space with RDP-compatible balanced sampling that accounts for group-specific noise levels across multiple sampling rates. In the latent space, we promote fairness by aligning neuron activation patterns across protected groups using Centered Kernel Alignment (CKA), a similarity measure extending the Hilbert-Schmidt Independence Criterion (HSIC). This alignment encourages statistical independence between latent representations and the protected feature. Empirical results demonstrate that FLIP effectively provides significant fairness improvements for task-agnostic fairness and across diverse downstream tasks under differential privacy constraints.

cs.LG↗

arXiv2025

Explanations as Bias Detectors: A Critical Study of Local Post-hoc XAI Methods for Fairness Exploration

As Artificial Intelligence (AI) is increasingly used in areas that significantly impact human lives, concerns about fairness and transparency have grown, especially regarding their impact on protected groups. Recently, the intersection of explainability and fairness has emerged as an important area to promote responsible AI systems. This paper explores how explainability methods can be leveraged to detect and interpret unfairness. We propose a pipeline that integrates local post-hoc explanation methods to derive fairness-related insights. During the pipeline design, we identify and address critical questions arising from the use of explanations as bias detectors such as the relationship between distributive and procedural fairness, the effect of removing the protected attribute, the consistency and quality of results across different explanation methods, the impact of various aggregation strategies of local explanations on group fairness evaluations, and the overall trustworthiness of explanations as bias detectors. Our results show the potential of explanation methods used for fairness while highlighting the need to carefully consider the aforementioned critical aspects.

cs.AI↗

arXiv2025

Achieving Socio-Economic Parity through the Lens of EU AI Act

Unfair treatment and discrimination are critical ethical concerns in AI systems, particularly as their adoption expands across diverse domains. Addressing these challenges, the recent introduction of the EU AI Act establishes a unified legal framework to ensure legal certainty for AI innovation and investment while safeguarding public interests, such as health, safety, fundamental rights, democracy, and the rule of law (Recital 8). The Act encourages stakeholders to initiate dialogue on existing AI fairness notions to address discriminatory outcomes of AI systems. However, these notions often overlook the critical role of Socio-Economic Status (SES), inadvertently perpetuating biases that favour the economically advantaged. This is concerning, given that principles of equalization advocate for equalizing resources or opportunities to mitigate disadvantages beyond an individual's control. While provisions for discrimination are laid down in the AI Act, specialized directions should be broadened, particularly in addressing economic disparities perpetuated by AI systems. In this work, we explore the limitations of popular AI fairness notions using a real-world dataset (Adult), highlighting their inability to address SES-driven disparities. To fill this gap, we propose a novel fairness notion, Socio-Economic Parity (SEP), which incorporates SES and promotes positive actions for underprivileged groups while accounting for factors within an individual's control, such as working hours, which can serve as a proxy for effort. We define a corresponding fairness measure and optimize a model constrained by SEP to demonstrate practical utility. Our results show the effectiveness of SEP in mitigating SES-driven biases. By analyzing the AI Act alongside our method, we lay a foundation for aligning AI fairness with SES factors while ensuring legal compliance.

cs.CY↗

arXiv2024

Emerging Security Challenges of Large Language Models

Large language models (LLMs) have achieved record adoption in a short period of time across many different sectors including high importance areas such as education [4] and healthcare [23]. LLMs are open-ended models trained on diverse data without being tailored for specific downstream tasks, enabling broad applicability across various domains. They are commonly used for text generation, but also widely used to assist with code generation [3], and even analysis of security information, as Microsoft Security Copilot demonstrates [18]. Traditional Machine Learning (ML) models are vulnerable to adversarial attacks [9]. So the concerns on the potential security implications of such wide scale adoption of LLMs have led to the creation of this working group on the security of LLMs. During the Dagstuhl seminar on "Network Attack Detection and Defense - AI-Powered Threats and Responses", the working group discussions focused on the vulnerability of LLMs to adversarial attacks, rather than their potential use in generating malware or enabling cyberattacks. Although we note the potential threat represented by the latter, the role of the LLMs in such uses is mostly as an accelerator for development, similar to what it is in benign use. To make the analysis more specific, the working group employed ChatGPT as a concrete example of an LLM and addressed the following points, which also form the structure of this report: 1. How do LLMs differ in vulnerabilities from traditional ML models? 2. What are the attack objectives in LLMs? 3. How complex it is to assess the risks posed by the vulnerabilities of LLMs? 4. What is the supply chain in LLMs, how data flow in and out of systems and what are the security implications? We conclude with an overview of open challenges and outlook.

cs.CR↗

arXiv2024

Shape error prediction in 5-axis machining using graph neural networks

This paper presents an innovative method for predicting shape errors in 5-axis machining using graph neural networks. The graph structure is defined with nodes representing workpiece surface points and edges denoting the neighboring relationships. The dataset encompasses data from a material removal simulation, process data, and post-machining quality information. Experimental results show that the presented approach can generalize the shape error prediction for the investigated workpiece geometry. Moreover, by modelling spatial and temporal connections within the workpiece, the approach handles a low number of labels compared to non-graphical methods such as Support Vector Machines.

eess.SY↗

arXiv2024

Unlocking LLMs: Addressing Scarce Data and Bias Challenges in Mental Health

Large language models (LLMs) have shown promising capabilities in healthcare analysis but face several challenges like hallucinations, parroting, and bias manifestation. These challenges are exacerbated in complex, sensitive, and low-resource domains. Therefore, in this work we introduce IC-AnnoMI, an expert-annotated motivational interviewing (MI) dataset built upon AnnoMI by generating in-context conversational dialogues leveraging LLMs, particularly ChatGPT. IC-AnnoMI employs targeted prompts accurately engineered through cues and tailored information, taking into account therapy style (empathy, reflection), contextual relevance, and false semantic change. Subsequently, the dialogues are annotated by experts, strictly adhering to the Motivational Interviewing Skills Code (MISC), focusing on both the psychological and linguistic dimensions of MI dialogues. We comprehensively evaluate the IC-AnnoMI dataset and ChatGPT's emotional reasoning ability and understanding of domain intricacies by modeling novel classification tasks employing several classical machine learning and current state-of-the-art transformer approaches. Finally, we discuss the effects of progressive prompting strategies and the impact of augmented data in mitigating the biases manifested in IC-AnnoM. Our contributions provide the MI community with not only a comprehensive dataset but also valuable insights for using LLMs in empathetic text generation for conversational therapy in supervised settings.

cs.CL↗

arXiv2024

Transparent Neighborhood Approximation for Text Classifier Explanation

Recent literature highlights the critical role of neighborhood construction in deriving model-agnostic explanations, with a growing trend toward deploying generative models to improve synthetic instance quality, especially for explaining text classifiers. These approaches overcome the challenges in neighborhood construction posed by the unstructured nature of texts, thereby improving the quality of explanations. However, the deployed generators are usually implemented via neural networks and lack inherent explainability, sparking arguments over the transparency of the explanation process itself. To address this limitation while preserving neighborhood quality, this paper introduces a probability-based editing method as an alternative to black-box text generators. This approach generates neighboring texts by implementing manipulations based on in-text contexts. Substituting the generator-based construction process with recursive probability-based editing, the resultant explanation method, XPROB (explainer with probability-based editing), exhibits competitive performance according to the evaluation conducted on two real-world datasets. Additionally, XPROB's fully transparent and more controllable construction process leads to superior stability compared to the generator-based explainers.

cs.CL↗

arXiv2024

Adversarial Robustness of VAEs across Intersectional Subgroups

Despite advancements in Autoencoders (AEs) for tasks like dimensionality reduction, representation learning and data generation, they remain vulnerable to adversarial attacks. Variational Autoencoders (VAEs), with their probabilistic approach to disentangling latent spaces, show stronger resistance to such perturbations compared to deterministic AEs; however, their resilience against adversarial inputs is still a concern. This study evaluates the robustness of VAEs against non-targeted adversarial attacks by optimizing minimal sample-specific perturbations to cause maximal damage across diverse demographic subgroups (combinations of age and gender). We investigate two questions: whether there are robustness disparities among subgroups, and what factors contribute to these disparities, such as data scarcity and representation entanglement. Our findings reveal that robustness disparities exist but are not always correlated with the size of the subgroup. By using downstream gender and age classifiers and examining latent embeddings, we highlight the vulnerability of subgroups like older women, who are prone to misclassification due to adversarial perturbations pushing their representations toward those of other subgroups.

cs.LG↗

arXiv2024

Fairness Evaluation with Item Response Theory

Item Response Theory (IRT) has been widely used in educational psychometrics to assess student ability, as well as the difficulty and discrimination of test questions. In this context, discrimination specifically refers to how effectively a question distinguishes between students of different ability levels, and it does not carry any connotation related to fairness. In recent years, IRT has been successfully used to evaluate the predictive performance of Machine Learning (ML) models, but this paper marks its first application in fairness evaluation. In this paper, we propose a novel Fair-IRT framework to evaluate a set of predictive models on a set of individuals, while simultaneously eliciting specific parameters, namely, the ability to make fair predictions (a feature of predictive models), as well as the discrimination and difficulty of individuals that affect the prediction results. Furthermore, we conduct a series of experiments to comprehensively understand the implications of these parameters for fairness evaluation. Detailed explanations for item characteristic curves (ICCs) are provided for particular individuals. We propose the flatness of ICCs to disentangle the unfairness between individuals and predictive models. The experiments demonstrate the effectiveness of this framework as a fairness evaluation tool. Two real-world case studies illustrate its potential application in evaluating fairness in both classification and regression tasks. Our paper aligns well with the Responsible Web track by proposing a Fair-IRT framework to evaluate fairness in ML models, which directly contributes to the development of a more inclusive, equitable, and trustworthy AI.

cs.CY↗

arXiv2024

TABCF: Counterfactual Explanations for Tabular Data Using a Transformer-Based VAE

In the field of Explainable AI (XAI), counterfactual (CF) explanations are one prominent method to interpret a black-box model by suggesting changes to the input that would alter a prediction. In real-world applications, the input is predominantly in tabular form and comprised of mixed data types and complex feature interdependencies. These unique data characteristics are difficult to model, and we empirically show that they lead to bias towards specific feature types when generating CFs. To overcome this issue, we introduce TABCF, a CF explanation method that leverages a transformer-based Variational Autoencoder (VAE) tailored for modeling tabular data. Our approach uses transformers to learn a continuous latent space and a novel Gumbel-Softmax detokenizer that enables precise categorical reconstruction while preserving end-to-end differentiability. Extensive quantitative evaluation on five financial datasets demonstrates that TABCF does not exhibit bias toward specific feature types, and outperforms existing methods in producing effective CFs that align with common CF desiderata.

cs.LG↗

arXiv2024

FairBranch: Mitigating Bias Transfer in Fair Multi-task Learning

The generalisation capacity of Multi-Task Learning (MTL) suffers when unrelated tasks negatively impact each other by updating shared parameters with conflicting gradients. This is known as negative transfer and leads to a drop in MTL accuracy compared to single-task learning (STL). Lately, there has been a growing focus on the fairness of MTL models, requiring the optimization of both accuracy and fairness for individual tasks. Analogously to negative transfer for accuracy, task-specific fairness considerations might adversely affect the fairness of other tasks when there is a conflict of fairness loss gradients between the jointly learned tasks - we refer to this as Bias Transfer. To address both negative- and bias-transfer in MTL, we propose a novel method called FairBranch, which branches the MTL model by assessing the similarity of learned parameters, thereby grouping related tasks to alleviate negative transfer. Moreover, it incorporates fairness loss gradient conflict correction between adjoining task-group branches to address bias transfer within these task groups. Our experiments on tabular and visual MTL problems show that FairBranch outperforms state-of-the-art MTLs on both fairness and accuracy.

cs.LG↗