arXiv subjects

Denis Parra

Publications and source records attributed to Denis Parra.

At least 19 recordsLinked to original sources

arXiv2026

Rethinking Layer Relevance in Large Language Models Beyond Cosine Similarity

Large language models (LLMs) have revolutionized natural language processing. Understanding their internal mechanisms is crucial for developing more interpretable and optimized architectures. Mechanistic interpretability has led to the development of various methods for assessing layer relevance, with cosine similarity being a widely used tool in the field. On this work, we demonstrate that cosine similarity is a poor proxy for the actual performance degradation caused by layer removal. Our theoretical analysis shows that a layer can exhibit an arbitrarily low cosine similarity score while still being crucial to the model's performance. On the other hand, empirical evidence from a range of LLMs confirms that the correlation between cosine similarity and actual performance degradation is often weak or moderate, leading to misleading interpretations of a transformer's internal mechanisms. We propose a more robust metric for assessing layer relevance: the actual drop in model accuracy resulting from the removal of a layer. Even though it is a computationally costly metric, this approach offers a more accurate picture of layer importance, allowing for more informed pruning strategies and lightweight models. Our findings have significant implications for the development of interpretable LLMs and highlight the need to move beyond cosine similarity in assessing layer relevance.

cs.LG↗

arXiv2026

Revisiting Framing Codebooks with AI: Employing Large Language Models as Analytical Collaborators in Deductive Content Analysis

Codebooks are central to framing research, providing theoretically grounded criteria for analyzing news content. While traditionally codebooks are built from theoretical frameworks and researchers' knowledge, applying these codebooks to large news corpora often exposes ambiguities, borderline cases, and underspecified rules that are difficult to resolve through theory alone. Moreover, news corpora evolve over time and differ across cultures, necessitating that researchers revisit the theoretical frameworks underlying these codebooks. In this article, we propose a workflow that uses Large Language Models (LLMs) to augment the creation and refinement of framing codebooks by combining theoretical frameworks with data-driven exploration. Rather than treating LLMs as automated classifiers, this approach positions them as analytic collaborators that help externalize decision rules, surface latent dimensions, and support iterative revisions of codebooks through dialogues between researchers and their data. We illustrate this workflow using a dataset of Latin American news coverage, demonstrating how the application of LLMs' capabilities has led to the surfacing of latent patterns, the generation of frame distinctions, and the adaptation of frameworks to new contexts. This method provides an LLM-assisted strategy that supports methodology creativity while preserving researchers' interpretative authority.

cs.HC↗

arXiv2026

ViT-Explainer: An Interactive Walkthrough of the Vision Transformer Pipeline

Transformer-based architectures have become the shared backbone of natural language processing and computer vision. However, understanding how these models operate remains challenging, particularly in vision settings, where images are processed as sequences of patch tokens. Existing interpretability tools often focus on isolated components or expert-oriented analysis, leaving a gap in guided, end-to-end understanding of the full inference pipeline. To bridge this gap, we present ViT-Explainer, a web-based interactive system that provides an integrated visualization of Vision Transformer inference, from patch tokenization to final classification. The system combines animated walkthroughs, patch-level attention overlays, and a vision-adapted Logit Lens within both guided and free exploration modes. A user study with six participants suggests that ViT-Explainer is easy to learn and use, helping users interpret and understand Vision Transformer behavior.

cs.CV↗

arXiv2026

Seeing to Generalize: How Visual Data Corrects Binding Shortcuts

Vision Language Models (VLMs) are designed to extend Large Language Models (LLMs) with visual capabilities, yet in this work we observe a surprising phenomenon: VLMs can outperform their underlying LLMs on purely text-only tasks, particularly in long-context information retrieval. To investigate this effect, we build a controlled synthetic retrieval task and find that a transformer trained only on text achieves perfect in-distribution accuracy but fails to generalize out of distribution, while subsequent training on an image-tokenized version of the same task nearly doubles text-only OOD performance. Mechanistic interpretability reveals that visual training changes the model's internal binding strategy: text-only training encourages positional shortcuts, whereas image-based training disrupts them through spatial translation invariance, forcing the model to adopt a more robust symbolic binding mechanism that persists even after text-only examples are reintroduced. We further characterize how binding strategies vary across training regimes, visual encoders, and initializations, and show that analogous shifts occur during pretrained LLM-to-VLM transitions. Our findings suggest that cross-modal training can enhance reasoning and generalization even for tasks grounded in a single modality.

cs.LG↗

arXiv2026

CURE: Curriculum-guided Multi-task Training for Reliable Anatomy Grounded Report Generation

Medical vision-language models can automate the generation of radiology reports but struggle with accurate visual grounding and factual consistency. Existing models often misalign textual findings with visual evidence, leading to unreliable or weakly grounded predictions. We present CURE, an error-aware curriculum learning framework that improves grounding and report quality without any additional data. CURE fine-tunes a multimodal instructional model on phrase grounding, grounded report generation, and anatomy-grounded report generation using public datasets. The method dynamically adjusts sampling based on model performance, emphasizing harder samples to improve spatial and textual alignment. CURE improves grounding accuracy by +0.35 IoU, boosts report quality by +0.192 CXRFEScore, and reduces hallucinations by 18.6%. CURE is a data-efficient framework that enhances both grounding accuracy and report reliability. Code is available at https://github.com/PabloMessina/CURE and model weights at https://huggingface.co/pamessina/medgemma-4b-it-cure

cs.CV↗

arXiv2025

Extending NGU to Multi-Agent RL: A Preliminary Study

The Never Give Up (NGU) algorithm has proven effective in reinforcement learning tasks with sparse rewards by combining episodic novelty and intrinsic motivation. In this work, we extend NGU to multi-agent environments and evaluate its performance in the simple_tag environment from the PettingZoo suite. Compared to a multi-agent DQN baseline, NGU achieves moderately higher returns and more stable learning dynamics. We investigate three design choices: (1) shared replay buffer versus individual replay buffers, (2) sharing episodic novelty among agents using different k thresholds, and (3) using heterogeneous values of the beta parameter. Our results show that NGU with a shared replay buffer yields the best performance and stability, highlighting that the gains come from combining NGU intrinsic exploration with experience sharing. Novelty sharing performs comparably when k = 1 but degrades learning for larger values. Finally, heterogeneous beta values do not improve over a small common value. These findings suggest that NGU can be effectively applied in multi-agent settings when experiences are shared and intrinsic exploration signals are carefully tuned.

cs.AI↗

arXiv2025

Perceptual Evaluation of GANs and Diffusion Models for Generating X-rays

Generative image models have achieved remarkable progress in both natural and medical imaging. In the medical context, these techniques offer a potential solution to data scarcity-especially for low-prevalence anomalies that impair the performance of AI-driven diagnostic and segmentation tools. However, questions remain regarding the fidelity and clinical utility of synthetic images, since poor generation quality can undermine model generalizability and trust. In this study, we evaluate the effectiveness of state-of-the-art generative models-Generative Adversarial Networks (GANs) and Diffusion Models (DMs)-for synthesizing chest X-rays conditioned on four abnormalities: Atelectasis (AT), Lung Opacity (LO), Pleural Effusion (PE), and Enlarged Cardiac Silhouette (ECS). Using a benchmark composed of real images from the MIMIC-CXR dataset and synthetic images from both GANs and DMs, we conducted a reader study with three radiologists of varied experience. Participants were asked to distinguish real from synthetic images and assess the consistency between visual features and the target abnormality. Our results show that while DMs generate more visually realistic images overall, GANs can report better accuracy for specific conditions, such as absence of ECS. We further identify visual cues radiologists use to detect synthetic images, offering insights into the perceptual gaps in current models. These findings underscore the complementary strengths of GANs and DMs and point to the need for further refinement to ensure generative models can reliably augment training datasets for AI diagnostic systems.

cs.CV↗

arXiv2025

User Perception of Attention Visualizations: Effects on Interpretability Across Evidence-Based Medical Documents

The attention mechanism is a core component of the Transformer architecture. Beyond improving performance, attention has been proposed as a mechanism for explainability via attention weights, which are associated with input features (e.g., tokens in a document). In this context, larger attention weights may imply more relevant features for the model's prediction. In evidence-based medicine, such explanations could support physicians' understanding and interaction with AI systems used to categorize biomedical literature. However, there is still no consensus on whether attention weights provide helpful explanations. Moreover, little research has explored how visualizing attention affects its usefulness as an explanation aid. To bridge this gap, we conducted a user study to evaluate whether attention-based explanations support users in biomedical document classification and whether there is a preferred way to visualize them. The study involved medical experts from various disciplines who classified articles based on study design (e.g., systematic reviews, broad synthesis, randomized and non-randomized trials). Our findings show that the Transformer model (XLNet) classified documents accurately; however, the attention weights were not perceived as particularly helpful for explaining the predictions. However, this perception varied significantly depending on how attention was visualized. Contrary to Munzner's principle of visual effectiveness, which favors precise encodings like bar length, users preferred more intuitive formats, such as text brightness or background color. While our results do not confirm the overall utility of attention weights for explanation, they suggest that their perceived helpfulness is influenced by how they are visually presented.

cs.CL↗

arXiv2025

A Systematic Review of User-Centred Evaluation of Explainable AI in Healthcare

Despite promising developments in Explainable Artificial Intelligence, the practical value of XAI methods remains under-explored and insufficiently validated in real-world settings. Robust and context-aware evaluation is essential, not only to produce understandable explanations but also to ensure their trustworthiness and usability for intended users, but tends to be overlooked because of no clear guidelines on how to design an evaluation with users. This study addresses this gap with two main goals: (1) to develop a framework of well-defined, atomic properties that characterise the user experience of XAI in healthcare; and (2) to provide clear, context-sensitive guidelines for defining evaluation strategies based on system characteristics. We conducted a systematic review of 82 user studies, sourced from five databases, all situated within healthcare settings and focused on evaluating AI-generated explanations. The analysis was guided by a predefined coding scheme informed by an existing evaluation framework, complemented by inductive codes developed iteratively. The review yields three key contributions: (1) a synthesis of current evaluation practices, highlighting a growing focus on human-centred approaches in healthcare XAI; (2) insights into the interrelations among explanation properties; and (3) an updated framework and a set of actionable guidelines to support interdisciplinary teams in designing and implementing effective evaluation strategies for XAI systems tailored to specific application contexts.

cs.HC↗

arXiv2025

CXR-LT 2024: A MICCAI challenge on long-tailed, multi-label, and zero-shot disease classification from chest X-ray

The CXR-LT series is a community-driven initiative designed to enhance lung disease classification using chest X-rays (CXR). It tackles challenges in open long-tailed lung disease classification and enhances the measurability of state-of-the-art techniques. The first event, CXR-LT 2023, aimed to achieve these goals by providing high-quality benchmark CXR data for model development and conducting comprehensive evaluations to identify ongoing issues impacting lung disease classification performance. Building on the success of CXR-LT 2023, the CXR-LT 2024 expands the dataset to 377,110 chest X-rays (CXRs) and 45 disease labels, including 19 new rare disease findings. It also introduces a new focus on zero-shot learning to address limitations identified in the previous event. Specifically, CXR-LT 2024 features three tasks: (i) long-tailed classification on a large, noisy test set, (ii) long-tailed classification on a manually annotated "gold standard" subset, and (iii) zero-shot generalization to five previously unseen disease findings. This paper provides an overview of CXR-LT 2024, detailing the data curation process and consolidating state-of-the-art solutions, including the use of multimodal models for rare disease detection, advanced generative approaches to handle noisy labels, and zero-shot learning strategies for unseen diseases. Additionally, the expanded dataset enhances disease coverage to better represent real-world clinical settings, offering a valuable resource for future research. By synthesizing the insights and innovations of participating teams, we aim to advance the development of clinically realistic and generalizable diagnostic models for chest radiography.

cs.CV↗

arXiv2025

The Role of Organizations in Networked Mobilization: Examining the 2011 Chilean Student Movement Through The Logic of Connective Action

This study examines the communication mechanisms that shape the formation of digitally-enabled mobilization networks. Informed by the logic of connective action, we postulate that the emergence of networks enabled by organizations and individuals is differentiated by network and framing mechanisms. From a case comparison within two mobilization networks -- one crowd-enabled and one organizationally-enabled -- of the 2011 Chilean student movement, we analyze their network structures and users' communication roles. We found that organizationally-enabled networks are likely to form from hierarchical cascades and crowd-enabled networks are likely to form from triadic closure mechanisms. Moreover, we found that organizations are essential for both kinds of networks: compared to individuals, organizations spread more messages among unconnected users, and organizations' messages are more likely to be spread. We discuss our findings in light of the network mechanisms and participation of organizations and influential users.

cs.SI↗

arXiv2025

A Compressive-Expressive Communication Framework for Compositional Representations

Compositionality in knowledge and language--the ability to represent complex concepts as a combination of simpler ones--is a hallmark of human cognition and communication. Despite recent advances, deep neural networks still struggle to acquire this property reliably. Neural models for emergent communication look to endow artificial agents with compositional language by simulating the pressures that form human language. In this work, we introduce CELEBI (Compressive-Expressive Language Emergence through a discrete Bottleneck and Iterated learning), a novel self-supervised framework for inducing compositional representations through a reconstruction-based communication game between a sender and a receiver. Building on theories of language emergence and the iterated learning framework, we integrate three mechanisms that jointly promote compressibility, expressivity, and efficiency in the emergent language. First, Progressive Decoding incentivizes intermediate reasoning by requiring the receiver to produce partial reconstructions after each symbol. Second, Final-State Imitation trains successive generations of agents to imitate reconstructions rather than messages, enforcing a tighter communication bottleneck. Third, Pairwise Distance Maximization regularizes message diversity by encouraging high distances between messages, with formal links to entropy maximization. Our method significantly improves both the efficiency and compositionality of the learned messages on the Shapes3D and MPI3D datasets, surpassing prior discrete communication frameworks in both reconstruction accuracy and topographic similarity. This work provides new theoretical and empirical evidence for the emergence of structured, generalizable communication protocols from simplicity-based inductive biases.

cs.LG↗

arXiv2024

Extracting and Encoding: Leveraging Large Language Models and Medical Knowledge to Enhance Radiological Text Representation

Advancing representation learning in specialized fields like medicine remains challenging due to the scarcity of expert annotations for text and images. To tackle this issue, we present a novel two-stage framework designed to extract high-quality factual statements from free-text radiology reports in order to improve the representations of text encoders and, consequently, their performance on various downstream tasks. In the first stage, we propose a \textit{Fact Extractor} that leverages large language models (LLMs) to identify factual statements from well-curated domain-specific datasets. In the second stage, we introduce a \textit{Fact Encoder} (CXRFE) based on a BERT model fine-tuned with objective functions designed to improve its representations using the extracted factual data. Our framework also includes a new embedding-based metric (CXRFEScore) for evaluating chest X-ray text generation systems, leveraging both stages of our approach. Extensive evaluations show that our fact extractor and encoder outperform current state-of-the-art methods in tasks such as sentence ranking, natural language inference, and label extraction from radiology reports. Additionally, our metric proves to be more robust and effective than existing metrics commonly used in the radiology report generation literature. The code of this project is available at \url{https://github.com/PabloMessina/CXR-Fact-Encoder}.

cs.CL↗

arXiv2024

Long Tail Image Generation Through Feature Space Augmentation and Iterated Learning

Image and multimodal machine learning tasks are very challenging to solve in the case of poorly distributed data. In particular, data availability and privacy restrictions exacerbate these hurdles in the medical domain. The state of the art in image generation quality is held by Latent Diffusion models, making them prime candidates for tackling this problem. However, a few key issues still need to be solved, such as the difficulty in generating data from under-represented classes and a slow inference process. To mitigate these issues, we propose a new method for image augmentation in long-tailed data based on leveraging the rich latent space of pre-trained Stable Diffusion Models. We create a modified separable latent space to mix head and tail class examples. We build this space via Iterated Learning of underlying sparsified embeddings, which we apply to task-specific saliency maps via a K-NN approach. Code is available at https://github.com/SugarFreeManatee/Feature-Space-Augmentation-and-Iterated-Learning

cs.CV↗

arXiv2023

Towards a Comprehensive Human-Centred Evaluation Framework for Explainable AI

While research on explainable AI (XAI) is booming and explanation techniques have proven promising in many application domains, standardised human-centred evaluation procedures are still missing. In addition, current evaluation procedures do not assess XAI methods holistically in the sense that they do not treat explanations' effects on humans as a complex user experience. To tackle this challenge, we propose to adapt the User-Centric Evaluation Framework used in recommender systems: we integrate explanation aspects, summarise explanation properties, indicate relations between them, and categorise metrics that measure these properties. With this comprehensive evaluation framework, we hope to contribute to the human-centred standardisation of XAI evaluation.

cs.HC↗

arXiv2021

Stress Test Evaluation of Biomedical Word Embeddings

The success of pretrained word embeddings has motivated their use in the biomedical domain, with contextualized embeddings yielding remarkable results in several biomedical NLP tasks. However, there is a lack of research on quantifying their behavior under severe "stress" scenarios. In this work, we systematically evaluate three language models with adversarial examples -- automatically constructed tests that allow us to examine how robust the models are. We propose two types of stress scenarios focused on the biomedical named entity recognition (NER) task, one inspired by spelling errors and another based on the use of synonyms for medical terms. Our experiments with three benchmarks show that the performance of the original models decreases considerably, in addition to revealing their weaknesses and strengths. Finally, we show that adversarial training causes the models to improve their robustness and even to exceed the original performance in some cases.

cs.CL↗

arXiv2021

Graphing else matters: exploiting aspect opinions and ratings in explainable graph-based recommendations

The success of neural network embeddings has entailed a renewed interest in using knowledge graphs for a wide variety of machine learning and information retrieval tasks. In particular, current recommendation methods based on graph embeddings have shown state-of-the-art performance. These methods commonly encode latent rating patterns and content features. Different from previous work, in this paper, we propose to exploit embeddings extracted from graphs that combine information from ratings and aspect-based opinions expressed in textual reviews. We then adapt and evaluate state-of-the-art graph embedding techniques over graphs generated from Amazon and Yelp reviews on six domains, outperforming baseline recommenders. Our approach has the advantage of providing explanations which leverage aspect-based opinions given by users about recommended items. Furthermore, we also provide examples of the applicability of recommendations utilizing aspect opinions as explanations in a visualization dashboard, which allows obtaining information about the most and least liked aspects of similar users obtained from the embeddings of an input graph.

cs.IR↗

arXiv2021

AHMoSe: A Knowledge-Based Visual Support System for Selecting Regression Machine Learning Models

Decision support systems have become increasingly popular in the domain of agriculture. With the development of automated machine learning, agricultural experts are now able to train, evaluate and make predictions using cutting edge machine learning (ML) models without the need for much ML knowledge. Although this automated approach has led to successful results in many scenarios, in certain cases (e.g., when few labeled datasets are available) choosing among different models with similar performance metrics is a difficult task. Furthermore, these systems do not commonly allow users to incorporate their domain knowledge that could facilitate the task of model selection, and to gain insight into the prediction system for eventual decision making. To address these issues, in this paper we present AHMoSe, a visual support system that allows domain experts to better understand, diagnose and compare different regression models, primarily by enriching model-agnostic explanations with domain knowledge. To validate AHMoSe, we describe a use case scenario in the viticulture domain, grape quality prediction, where the system enables users to diagnose and select prediction models that perform better. We also discuss feedback concerning the design of the tool from both ML and viticulture experts.

cs.HC↗