Search

Many real-world eligibility problems, ranging from medical diagnosis to tax planning, can be mapped to decision problems expressed in natural language, wherein a model must make a binary choice based on user features. Large-scale domains such as legal codes or frequently updated funding opportunities render human annotation (e.g., web forms or decision trees) impractical, highlighting the need for agents that can automatically assist in decision-making. Since relevant information is often only known to the user, it is crucial that these agents ask the right questions. As agents determine when to terminate a conversation, they face a trade-off between accuracy and the number of questions asked, a key metric for both user experience and cost. To evaluate this task, we propose BeNYfits, a new benchmark for determining user eligibility for multiple overlapping social benefits opportunities through interactive decision-making. Our experiments show that current language models struggle with frequent hallucinations, with GPT-4o scoring only 35.7 F1 using a ReAct-style chain-of-thought. To address this, we introduce ProADA, a novel approach that leverages program synthesis to assist in decision-making by mapping dialog planning to a code generation problem and using gaps in structured data to determine the best next action. Our agent, ProADA, improves the F1 score to 55.6 while maintaining nearly the same number of dialog turns.

There has been significant research on propagandistic content detection across different modalities and languages. However, most studies have primarily focused on detection, with little attention given to explanations justifying the predicted label. This is largely due to the lack of resources that provide explanations alongside annotated labels. To address this issue, we propose a multilingual (i.e., Arabic and English) explanation-enhanced dataset, the first of its kind. Additionally, we introduce an explanation-enhanced LLM for both label detection and rationale-based explanation generation. Our findings indicate that the model performs comparably while also generating explanations. We will make the dataset and experimental resources publicly available for the research community.

We introduce a novel multilingual hierarchical corpus annotated for entity framing and role portrayal in news articles. The dataset uses a unique taxonomy inspired by storytelling elements, comprising 22 fine-grained roles, or archetypes, nested within three main categories: protagonist, antagonist, and innocent. Each archetype is carefully defined, capturing nuanced portrayals of entities such as guardian, martyr, and underdog for protagonists; tyrant, deceiver, and bigot for antagonists; and victim, scapegoat, and exploited for innocents. The dataset includes 1,378 recent news articles in five languages (Bulgarian, English, Hindi, European Portuguese, and Russian) focusing on two critical domains of global significance: the Ukraine-Russia War and Climate Change. Over 5,800 entity mentions have been annotated with role labels. This dataset serves as a valuable resource for research into role portrayal and has broader implications for news analysis. We describe the characteristics of the dataset and the annotation process, and we report evaluation results on fine-tuned state-of-the-art multilingual transformers and hierarchical zero-shot learning using LLMs at the level of a document, a paragraph, and a sentence., Comment: 23 pages, 12 figures. Submitted to ACL Rolling Review (ARR)

The integration of Large Language Models (LLMs) into optimization has created a powerful synergy, opening exciting research opportunities. This paper investigates how LLMs can enhance existing optimization algorithms. Using their pre-trained knowledge, we demonstrate their ability to propose innovative heuristic variations and implementation strategies. To evaluate this, we applied a non-trivial optimization algorithm, Construct, Merge, Solve and Adapt (CMSA) -- a hybrid metaheuristic for combinatorial optimization problems that incorporates a heuristic in the solution construction phase. Our results show that an alternative heuristic proposed by GPT-4o outperforms the expert-designed heuristic of CMSA, with the performance gap widening on larger and denser graphs. Project URL: https://imp-opt-algo-llms.surge.sh/

Embedded Field-Programmable Gate Arrays (eFPGAs) allow for the design of hardware accelerators of edge Machine Learning (ML) applications at a lower power budget compared with traditional FPGA platforms. However, the limited eFPGA logic and memory significantly constrain compute capabilities and model size. As such, ML application deployment on eFPGAs is in direct contrast with the most recent FPGA approaches developing architecture-specific implementations and maximizing throughput over resource frugality. This paper focuses on the opposite side of this trade-off: the proposed eFPGA accelerator focuses on minimizing resource usage and allowing flexibility for on-field recalibration over throughput. This allows for runtime changes in model size, architecture, and input data dimensionality without offline resynthesis. This is made possible through the use of a bitwise compressed inference architecture of the Tsetlin Machine (TM) algorithm. TM compute does not require any multiplication operations, being limited to only bitwise AND, OR, NOT, summations and additions. Additionally, TM model compression allows the entire model to fit within the on-chip block RAM of the eFPGA. The paper uses this accelerator to propose a strategy for runtime model tuning in the field. The proposed approach uses 2.5x fewer Look-up-Tables (LUTs) and 3.38x fewer registers than the current most resource-fugal design and achieves up to 129x energy reduction compared with low-power microcontrollers running the same ML application., Comment: Accepted as a full paper by the 2025 EDGE AI FOUNDATION Austin

Social media networks have amplified the reach of social and political movements, but most research focuses on mainstream platforms such as X, Reddit, and Facebook, overlooking Discord. As a rapidly growing, community-driven platform with optional decentralized moderation, Discord offers unique opportunities to study political discourse. This study analyzes over 30 million messages from political servers on Discord discussing the 2024 U.S. elections. Servers were classified as Republican-aligned, Democratic-aligned, or unaligned based on their descriptions. We tracked changes in political conversation during key campaign events and identified distinct political valence and implicit biases in semantic association through embedding analysis. We observed that Republican servers emphasized economic policies, while Democratic servers focused on equality-related and progressive causes. Furthermore, we detected an increase in toxic language, such as sexism, in Republican-aligned servers after Kamala Harris's nomination. These findings provide a first look at political behavior on Discord, highlighting its growing role in shaping and understanding online political engagement.

Survival analysis, a foundational tool for modeling time-to-event data, has seen growing integration with machine learning (ML) approaches to handle the complexities of censored data and time-varying risks. Despite these advances, leveraging state-of-the-art survival models remains a challenge due to the fragmented nature of existing implementations, which lack standardized interfaces and require extensive preprocessing. We introduce SurvHive, a Python-based framework designed to unify survival analysis methods within a coherent and extensible interface modeled on scikit-learn. SurvHive integrates classical statistical models with cutting-edge deep learning approaches, including transformer-based architectures and parametric survival models. Using a consistent API, SurvHive simplifies model training, evaluation, and optimization, significantly reducing the barrier to entry for ML practitioners exploring survival analysis. The package includes enhanced support for hyper-parameter tuning, time-dependent risk evaluation metrics, and cross-validation strategies tailored to censored data. With its extensibility and focus on usability, SurvHive provides a bridge between survival analysis and the broader ML community, facilitating advancements in time-to-event modeling across domains. The SurvHive code and documentation are available freely at https://github.com/compbiomed-unito/survhive., Comment: 11 loages, 1 table

The cross-correlation of cosmic voids with the lensing convergence ($\kappa$) map of the Cosmic Microwave Background (CMB) fluctuations provides a powerful tool to refine our understanding of the cosmological model. However, several studies have reported a moderate tension between the lensing imprint of cosmic voids on the observed CMB and the simulated $\mathrm{\Lambda}$CDM signal. To address this "lensing-is-low" tension and to obtain new, precise measurements, we exploit the large DESI Legacy Survey Luminous Red Galaxy (LRG) dataset, covering approximately 19,500 $\deg^2$ of the sky and including about 10 million LRGs at $z < 1.05$. Our $\mathrm{\Lambda}$CDM template was created using the Buzzard mocks, which we specifically calibrated to match the clustering properties of the observed galaxy sample by exploiting more than one million DESI spectra. We identified our catalogs of 3D voids in the range $0.35 < z < 0.95$, dividing the sample into bins according to the redshift and $\lambda_\mathrm{v}$ values of the voids. We report a 14$\sigma$ detection of the lensing signal, with $A_\kappa = 1.016 \pm 0.054$, which increases to 17$\sigma$ when considering the void-in-void ($A_\kappa = 0.944 \pm 0.064$) and the void-in-cloud ($A_\kappa = 0.975 \pm 0.060$) populations individually, the highest detection significance for studies of this kind. We observe a full agreement between the observations and $\mathrm{\Lambda}$CDM predictions across all redshift bins, sky regions, and void populations considered. In addition to these findings, our analysis highlights the importance of matching sparseness and redshift error distributions between mocks and observations, as well as the role of $\lambda_\mathrm{v}$ in enhancing the signal-to-noise ratio., Comment: 19 pages, 13 figures

The fast advancement of Large Vision-Language Models (LVLMs) has shown immense potential. These models are increasingly capable of tackling abstract visual tasks. Geometric structures, particularly graphs with their inherent flexibility and complexity, serve as an excellent benchmark for evaluating these models' predictive capabilities. While human observers can readily identify subtle visual details and perform accurate analyses, our investigation reveals that state-of-the-art LVLMs exhibit consistent limitations in specific visual graph scenarios, especially when confronted with stylistic variations. In response to these challenges, we introduce VisGraphVar (Visual Graph Variability), a customizable benchmark generator able to produce graph images for seven distinct task categories (detection, classification, segmentation, pattern recognition, link prediction, reasoning, matching), designed to systematically evaluate the strengths and limitations of individual LVLMs. We use VisGraphVar to produce 990 graph images and evaluate six LVLMs, employing two distinct prompting strategies, namely zero-shot and chain-of-thought. The findings demonstrate that variations in visual attributes of images (e.g., node labeling and layout) and the deliberate inclusion of visual imperfections, such as overlapping nodes, significantly affect model performance. This research emphasizes the importance of a comprehensive evaluation across graph-related tasks, extending beyond reasoning alone. VisGraphVar offers valuable insights to guide the development of more reliable and robust systems capable of performing advanced visual graph analysis.

This paper addresses the problem of learning the impulse responses characterizing forward models by means of a regularized kernel-based Prediction Error Method (PEM). The common approach to accomplish that is to approximate the system with a high-order stable ARX model. However, such choice induces a certain undesired prior information in the system that we want to estimate. To overcome this issue, we propose a new kernel-based paradigm which is formulated directly in terms of the impulse responses of the forward model and leading to the identification of a high-order MAX model. The most challenging step is the estimation of the kernel hyperparameters optimizing the marginal likelihood. The latter, indeed, does not admit a closed form expression. We propose a method for evaluating the marginal likelihood which makes possible the hyperparameters estimation. Finally, some numerical results showing the effectiveness of the method are presented.

We present a modular toolkit to perform joint speaker diarization and speaker identification. The toolkit can leverage on multiple models and algorithms which are defined in a configuration file. Such flexibility allows our system to work properly in various conditions (e.g., multiple registered speakers' sets, acoustic conditions and languages) and across application domains (e.g. media monitoring, institutional, speech analytics). In this demonstration we show a practical use-case in which speaker-related information is used jointly with automatic speech recognition engines to generate speaker-attributed transcriptions. To achieve that, we employ a user-friendly web-based interface to process audio and video inputs with the chosen configuration., Comment: Show and Tell paper. Presented at Interspeech 2024

The Exame Nacional do Ensino M\'edio (ENEM) is a pivotal test for Brazilian students, required for admission to a significant number of universities in Brazil. The test consists of four objective high-school level tests on Math, Humanities, Natural Sciences and Languages, and one writing essay. Students' answers to the test and to the accompanying socioeconomic status questionnaire are made public every year (albeit anonymized) due to transparency policies from the Brazilian Government. In the context of large language models (LLMs), these data lend themselves nicely to comparing different groups of humans with AI, as we can have access to human and machine answer distributions. We leverage these characteristics of the ENEM dataset and compare GPT-3.5 and 4, and MariTalk, a model trained using Portuguese data, to humans, aiming to ascertain how their answers relate to real societal groups and what that may reveal about the model biases. We divide the human groups by using socioeconomic status (SES), and compare their answer distribution with LLMs for each question and for the essay. We find no significant biases when comparing LLM performance to humans on the multiple-choice Brazilian Portuguese tests, as the distance between model and human answers is mostly determined by the human accuracy. A similar conclusion is found by looking at the generated text as, when analyzing the essays, we observe that human and LLM essays differ in a few key factors, one being the choice of words where model essays were easily separable from human ones. The texts also differ syntactically, with LLM generated essays exhibiting, on average, smaller sentences and less thought units, among other differences. These results suggest that, for Brazilian Portuguese in the ENEM context, LLM outputs represent no group of humans, being significantly different from the answers from Brazilian students across all tests., Comment: Accepted at the Seventh AAAI/ACM Conference on AI, Ethics and Society (AIES 2024). 14 pages, 4 figures

The Southern Photometric Local Universe Survey (S-PLUS) is a project to map $\sim9300$ sq deg of the sky using twelve bands (seven narrow and five broadbands). Observations are performed with the T80-South telescope, a robotic telescope located at the Cerro Tololo Observatory in Chile. The survey footprint consists of several large contiguous areas, including fields at high and low galactic latitudes, and towards the Magellanic Clouds. S-PLUS uses fixed exposure times to reach point source depths of about $21$ mag in the $griz$ and $20$ mag in the $u$ and the narrow filters. This paper describes the S-PLUS Data Release 4 (DR4), which includes calibrated images and derived catalogues for over 3000 sq deg, covering the aforementioned area. The catalogues provide multi-band photometry performed with the tools \texttt{DoPHOT} and \texttt{SExtractor} -- point spread function (\PSF) and aperture photometry, respectively. In addition to the characterization, we also present the scientific potential of the data. We use statistical tools to present and compare the photometry obtained through different methods. Overall we find good agreement between the different methods, with a slight systematic offset of 0.05\,mag between our \PSF and aperture photometry. We show that the astrometry accuracy is equivalent to that obtained in previous S-PLUS data releases, even in very crowded fields where photometric extraction is challenging. The depths of main survey (MS) photometry for a minimum signal-to-noise ratio $S/N = 3$ reach from $\sim19.5$ for the bluer bands to $\sim21.5$ mag on the red. The range of magnitudes over which accurate \PSF photometry is obtained is shallower, reaching $\sim19$ to $\sim20.5$ mag depending on the filter. Based on these photometric data, we provide star-galaxy-quasar classification and photometric redshift for millions of objects., Comment: 26 pages, 17 figures, 14 tables, accepted for A&A

Wearable assistive robots (WR) for the lower extremity are extensively documented in literature. Various interfaces have been designed to control these devices during gait and balance activities. However, achieving seamless and intuitive control requires accurate modeling of the human neuromusculoskeletal (NMSK) system. Such modeling enables WR to anticipate user intentions and determine the necessary joint assistance. Despite the existence of controllers interfacing with the NMSK system, robust and generalizable techniques across different tasks remain scarce. Designing these novel controllers necessitates the combined expertise of neurophysiologists, who understand the physiology of movement initiation and generation, and biomechatronic engineers, who design and control devices that assist movement. This paper aims to bridge the gaps between these fields by presenting a primer on key concepts and the current state of the science in each area. We present three main sections: the neuromechanics of locomotion, neuromechanical models of movement, and existing neuromechanical controllers used in WR. Through these sections, we provide a comprehensive overview of seminal studies in the field, facilitating collaboration between neurophysiologists and biomechatronic engineers for future advances in wearable robotics for locomotion.

Context. The dissipation process responsible for the long gamma-ray burst (GRB) prompt emission and the kind of dynamics that drives the release of energy as a function of time are still key open issues. We recently found that the distribution of the number of peaks per GRB is described by a mixture of two exponentials, suggesting the existence of two behaviours that turn up as peak-rich and peak-poor time profiles. Aims. Our aims are to study the distribution of the number of peaks per GRB of the entire catalogue of about 3000 GRBs observed by the Fermi Gamma-ray Burst Monitor (GBM) and to make a comparison with previous results obtained from other catalogues. Methods. We identified GRB peaks using the MEPSA code and modelled the resulting distribution following the same procedure that was adopted in the previous analogous investigation. Results. We confirm that only a mixture of two exponentials can model the distribution satisfactorily, with model parameters that fully agree with those found from previous analyses. In particular, we confirm that (21 +- 4)% of the observed GRBs are peak-rich (8 +- 1 peaks per GRB on average), while the remaining 80% are peak-poor (2.12 +- 0.10 peaks per GRB on average). Conclusions. We confirm the existence of two different components, peak-poor and peak-rich GRBs, that make up the observed GRB populations. Together with previous analogous results from other GRB catalogues, these results provide compelling evidence that GRB prompt emission is governed by two distinct regimes., Comment: 6 pages, 3 figures, accepted as a letter by A&A

Like the letters in the alphabet forming words, reusing components of a heterogeneous mixture is an efficient strategy for assembling a large number of target structures. Examples range from synthetic DNA origami to proteins self-assembling into complexes. The standard self-assembly paradigm views target structures as free-energy minima of a mixture. While this is an appealing picture, at high speed structures may be kinetically trapped in local minima, reducing self-assembly accuracy. How then can high speed, high accuracy, and combinatorial usage of components coexist? We propose to reconcile these three concepts not by avoiding kinetic traps, but by exploiting them to encode target structures. This can be achieved by sculpting the kinetic pathways of the mixture, instead of its free-energy landscape. We formalize these ideas in a minimal toy model, for which we analytically estimate the encoding capacity and kinetic characteristics, in agreement with simulations. Our results may be generalized to other soft-matter systems capable of computation, such as liquid mixtures or elastic networks, and pave the way for high-dimensional information processing far from equilibrium., Comment: Main: 6 pages, 4 figures. SI: 3 pages, 3 figures

Since the rise of Large Language Models (LLMs) a couple of years ago, researchers in metaheuristics (MHs) have wondered how to use their power in a beneficial way within their algorithms. This paper introduces a novel approach that leverages LLMs as pattern recognition tools to improve MHs. The resulting hybrid method, tested in the context of a social network-based combinatorial optimization problem, outperforms existing state-of-the-art approaches that combine machine learning with MHs regarding the obtained solution quality. By carefully designing prompts, we demonstrate that the output obtained from LLMs can be used as problem knowledge, leading to improved results. Lastly, we acknowledge LLMs' potential drawbacks and limitations and consider it essential to examine them to advance this type of research further. Our method can be reproduced using a tool available at: https://github.com/camilochs/optipattern.

The current standard model of cosmology successfully describes a variety of measurements, but the nature of its main ingredients, dark matter and dark energy, remains unknown. Euclid is a medium-class mission in the Cosmic Vision 2015-2025 programme of the European Space Agency (ESA) that will provide high-resolution optical imaging, as well as near-infrared imaging and spectroscopy, over about 14,000 deg^2 of extragalactic sky. In addition to accurate weak lensing and clustering measurements that probe structure formation over half of the age of the Universe, its primary probes for cosmology, these exquisite data will enable a wide range of science. This paper provides a high-level overview of the mission, summarising the survey characteristics, the various data-processing steps, and data products. We also highlight the main science objectives and expected performance., Comment: Accepted for publication in the A&A special issue`Euclid on Sky'

Quantum computing represents a cutting-edge frontier in artificial intelligence. It makes use of hybrid quantum-classical computation which tries to leverage quantum mechanic principles that allow us to use a different approach to deep learning classification problems. The work presented here falls within the context of the AGILE space mission, launched in 2007 by the Italian Space Agency. We implement different Quantum Convolutional Neural Networks (QCNN) that analyze data acquired by the instruments onboard AGILE to detect Gamma-Ray Bursts from sky maps or light curves. We use several frameworks such as TensorFlow-Quantum, Qiskit and PennyLane to simulate a quantum computer. We achieved an accuracy of 95.1% on sky maps with QCNNs, while the classical counterpart achieved 98.8% on the same data, using however hundreds of thousands more parameters., Comment: 4 pages, 2 figures, proceedings of the ADASS XXXIII (2023) conference, to appear in ASP Conference Serie

Modern Web APIs allow developers to provide extensively customized experiences for website visitors, but the richness of the device information they provide also make them vulnerable to being abused to construct browser fingerprints, device-specific identifiers that enable covert tracking of users even when cookies are disabled. Previous research has established entropy, a measure of information, as the key metric for quantifying fingerprinting risk. However, earlier studies had two major limitations. First, their entropy estimates were based on either a single website or a very small sample of devices. Second, they did not adequately consider correlations among different Web APIs, potentially grossly overestimating their fingerprinting risk. We provide the first study of browser fingerprinting which addresses the limitations of prior work. Our study is based on actual visited pages and Web APIs reported by tens of millions of real Chrome browsers in-the-wild. We accounted for the dependencies and correlations among Web APIs, which is crucial for obtaining more realistic entropy estimates. We also developed a novel experimental design that accurately and efficiently estimates entropy while never observing too much information from any single user. Our results provide an understanding of the distribution of entropy for different website categories, confirm the utility of entropy as a fingerprinting proxy, and offer a method for evaluating browser enhancements which are intended to mitigate fingerprinting., Comment: A version of this report to appear in the proceedings of The Web Conference (WWW) 2024. This version contains additional material in the appendix

Testing the equality of mean vectors across $g$ different groups plays an important role in many scientific fields. In regular frameworks, likelihood-based statistics under the normality assumption offer a general solution to this task. However, the accuracy of standard asymptotic results is not reliable when the dimension $p$ of the data is large relative to the sample size $n_i$ of each group. We propose here an exact directional test for the equality of $g$ normal mean vectors with identical unknown covariance matrix, provided that $\sum_{i=1}^g n_i \ge p+g+1$. In the case of two groups ($g=2$), the directional test is equivalent to the Hotelling's $T^2$ test. In the more general situation where the $g$ independent groups may have different unknown covariance matrices, although exactness does not hold, simulation studies show that the directional test is more accurate than most commonly used likelihood based solutions. Robustness of the directional approach and its competitors under deviation from multivariate normality is also numerically investigated., Comment: 51 pages, 15 figures

The (sub)millimeter sky contains a vast wealth of information that is both complementary and inaccessible to other wavelengths. Over half the light we receive is observable at (sub)millimeter wavelengths, yet we have mapped only a small portion of the sky at sufficient spatial resolution and sensitivity to detect and resolve distant galaxies or star forming cores within their large-scale environments. For decades the astronomical community has highlighted the need for a large, high-throughput (sub-)mm ($\lambda\sim 0.35-10$ mm) single dish. The Atacama Large Aperture Submillimeter Telescope (AtLAST), with its 50-m aperture and $2^\circ$ maximal field of view, aims to be such a facility. We present here the preliminary design concept for AtLAST, developed through an EU Horizon 2020-funded design study. Our design approach begins with a long lineage of (sub)millimeter telescopes, relies on calculations and simulations to realize the optics, and uses finite element analysis to optimize the designs for the mechanical structure and subsystems. The demanding technical requirements for AtLAST, set by transformative science goals, have motivated the design effort to combine novel concepts with lessons learned from the past experience of previous efforts. The result is an innovative rocking chair design with six instrument bays, two of which are mounted on Nasmyth platforms, inside a large receiver cabin. Ultimately, AtLAST aims to achieve a surface accuracy of $\leq 20~\mu$m root mean square half wavefront error, corresponding a Ruze efficiency $>50\%$ at 950~GHz. We conclude that closed-loop metrology of the active primary surface will likely be required to achieve our surface accuracy goal. In the next phase of the project, we will prototype and test such metrology on existing platforms, with a goal of delivering a mature, construction-ready design by the end of this decade., Comment: Accepted 2025-Jan-14. 28 pages, 22 figures, 6 tables. Supplementary materials that were previously in Appendix B & C can be found on Zenodo at https://zenodo.org/records/14673697

The variety of long duration gamma-ray burst (LGRB) light curves (LCs) encode a wealth of information on how LGRB engines release energy following the collapse of the progenitor star. Attempts to characterise GRB LCs focused on a number of properties, such as the minimum variability timescale, power density spectra (both ensemble average and individual), or with different definitions of variability. In parallel, a characterisation as a stochastic process was pursued by studying the distributions of waiting times, peak flux, fluence of individual peaks within GRB time profiles. Yet, the question remains as to whether the diversity of profiles can be described in terms of a common stochastic process. Here we address this issue by studying for the first time the distribution of the number of peaks in a GRB profile. We used four different GRB catalogues: CGRO/BATSE, Swift/BAT, BeppoSAX/GRBM, and Insight-HXMT. The statistically significant peaks were identified by means of well tested algorithm MEPSA and further selected by applying a set of thresholds on signal-to-noise ratio. We then extracted the corresponding distributions of number of peaks per GRB. Among the different models considered (power-law, simple or stretched exponential) only a mixture of two exponentials models all the observed distributions, suggesting the existence of two distinct behaviours: (i) an average number of 2.1+-0.1 peaks per GRB ("peak poor") and accounting for about 80% of the observed population of GRBs; (ii) an average number of 8.3+-1.0 peaks per GRB ("peak rich") and accounting for the remaining 20% of the observed population. We associate the class of peak-rich GRBs with the presence of sub-second variability, which seems to be absent among peak-poor GRBs. The two classes could result from two different regimes through which GRB engines release energy or through which energy is dissipated into gamma-rays., Comment: 7 pages, 5 figures, accepted by A&A

The ability of Large Language Models (LLMs) to generate high-quality text and code has fuelled their rise in popularity. In this paper, we aim to demonstrate the potential of LLMs within the realm of optimization algorithms by integrating them into STNWeb. This is a web-based tool for the generation of Search Trajectory Networks (STNs), which are visualizations of optimization algorithm behavior. Although visualizations produced by STNWeb can be very informative for algorithm designers, they often require a certain level of prior knowledge to be interpreted. In an attempt to bridge this knowledge gap, we have incorporated LLMs, specifically GPT-4, into STNWeb to produce extensive written reports, complemented by automatically generated plots, thereby enhancing the user experience and reducing the barriers to the adoption of this tool by the research community. Moreover, our approach can be expanded to other tools from the optimization community, showcasing the versatility and potential of LLMs in this field., Comment: Submitted to the GECCO 2024 conference