Archivio della ricerca - Fondazione Bruno Kessler
Not a member yet
21227 research outputs found
Sort by
Measurements of W+W− production cross-sections in pp collisions at √s = 13 TeV with the ATLAS detector
No full textMeasurements of W+W− → e±νμ∓ν production cross-sections are presented, providing a test of the predictions of perturbative quantum chromodynamics and the electroweak theory. The measurements are based on data from pp collisions at
= 13 TeV recorded by the ATLAS detector at the Large Hadron Collider in 2015–2018, corresponding to an integrated luminosity of 140 fb−1. The number of events due to top-quark pair production, the largest background, is reduced by rejecting events containing jets with b-hadron decays. An improved methodology for estimating the remaining top-quark background enables a precise measurement of W+W− cross-sections with no additional requirements on jets. The fiducial W+W− cross-section is determined in a maximum-likelihood fit with an uncertainty of 3.1%. The measurement is extrapolated to the full phase space, resulting in a total W+W− cross-section of 127 ± 4 pb. Differential cross-sections are measured as a function of twelve observables that comprehensively describe the kinematics of W+W− events. The measurements are compared with state-of-the-art theory calculations and excellent agreement with predictions is observed. A charge asymmetry in the lepton rapidity is observed as a function of the dilepton invariant mass, in agreement with the Standard Model expectation. A CP-odd observable is measured to be consistent with no CP violation. Limits on Standard Model effective field theory Wilson coefficients in the Warsaw basis are obtained from the differential cross-sections
Beauty and the bias: Exploring the impact of attractiveness on multimodal large language models
No full textPhysical attractiveness matters. It has been shown to influence human perception and decision-making, often leading to biased judgments that favor those deemed attractive in what is referred to as the "attractiveness halo effect". While extensively studied in human judgments in a broad set of domains, including hiring, judicial sentencing or credit granting, the role that attractiveness plays in the assessments and decisions made by multimodal large language models (MLLMs) is unknown. To address this gap, we conduct an empirical study with 7 diverse open-source MLLMs evaluated on 91 socially relevant scenarios and a diverse dataset of 924 face images --corresponding to 462 individuals both with and without beauty filters applied to them. Our analysis reveals that attractiveness impacts the decisions made by MLLMs in 86.2% of the scenarios on average, demonstrating substantial bias in model behavior in what we refer to as an attractiveness bias. Similarly to humans, we find empirical evidence of the existence of the attractiveness halo effect in 94.8% of the relevant scenarios: attractive individuals are more likely to be attributed positive traits, such as trustworthiness or confidence, by MLLMs than unattractive individuals. Furthermore, we uncover gender, age and race biases in a significant portion of the scenarios which are also impacted by attractiveness, particularly in the case of gender, highlighting the intersectional nature of the algorithmic attractiveness bias. Our findings suggest that societal stereotypes and cultural norms intersect with perceptions of attractiveness in MLLMs in a complex manner. Our work emphasizes the need to account for intersectionality in algorithmic bias detection and mitigation efforts and underscores the challenges of addressing biases in modern MLLMs
Mapping B cells and the immune landscape of tertiary lymphoid structures reveals their clinical impact in neuroblastoma
No full textBackground: Immunotherapy has transformed cancer treatment, highlighting the importance of effective antitumor immunity to fight cancer. However, its success in pediatric cancer remains limited, underscoring the urgent need to identify new immunotherapeutic targets. In this study, we explored the clinical relevance of B cells and tertiary lymphoid structures (TLS) in neuroblastoma (NB), a pediatric tumor with a heterogeneous immune landscape. Methods: We analyzed 87 treatment-naïve NB specimens, spanning both localized and metastatic disease previously characterized for T-cell and dendritic cell (DC) infiltration. B cells were detected by immunohistochemistry, and plasma cells were quantified using multiple immunofluorescence. Spatial organization and functional status of immune cells within TLSs were assessed by imaging mass cytometry using a 29-antibody panel. In parallel, gene expression profiles were obtained through NanoString PanCancer Immune Profiling and further validated using publicly available bulk and single-cell RNA-sequencing data from untreated and treated NB samples. These transcriptomic datasets were used to support protein-level findings and to identify prognostic gene signatures. Results: B-cell infiltration in NB tumors strongly correlated with the presence of T cells and DCs at both protein and transcriptomic levels, and was associated with improved prognosis. Similar to other solid tumors, B cells in NB were either scattered throughout the tumor or organized into TLSs of varying maturity. Spatial proteomic and transcriptomic analyses revealed that localized tumors often contain mature TLSs, with functional B cells able to antigen presentation and immunoglobulin expression, alongside high cytotoxic T cells. In contrast, metastatic tumors primarily exhibited immature TLSs, with evidence of B-cell and T-cell dysfunction. Importantly, we identified gene signatures associated with B cells and TLSs that not only predicted survival in NB but were also prognostic in multiple adult cancers. Conclusions: Our findings highlight a central role for B cells and TLSs in shaping the immune microenvironment of NB. Their presence and maturation status are linked to clinical outcome, suggesting their potential as prognostic biomarkers and targets for novel immunotherapeutic strategies in pediatric oncology
Simulation and Characterization of a Monolithic Active Pixel Sensor Designed for Brachytherapy
No full textIn order to be effective and to reduce potential issues related to late toxicities in the healthy tissues, Low Dose Rate (LDR) and High Dose Rate (HDR) brachytherapy treatments require both an excellent control on the actual position of the radioactive seeds implanted within the patient body to deliver the intended dose to the target cancer cells. To this end, we developed a small area pad detector (O (1 mm2)) with embedded CMOS electronics that can be mounted on the tip of a thin and flexible prostate catheter. This results in a compact system that can be exploited to monitor the clinical treatment performing in-vivo dosimetry and source tracking, thus improving the therapy outcome through the prompt identification of potential errors related to the misplacement of the implanted radioactive seeds. To assess the detector performance, we electrically characterized the sensor monitoring the chip power consumption. Then, we tested its dynamic response using an optical setup equipped with a fast pulsed IR laser with 1060 nm wavelength coupled to a focusing system able to provide a minimum laser spot size in the order of 10 μm FWHM and we compared the obtained waveforms with the ones predicted by simulations. Finally, we evaluated the detector response to an X-ray tube and different radiation sources, e.g. 90Sr and 133Ba, to calibrate the gain of the embedded amplifier and we exploited an external charge injection circuit to verify the linear dynamic range of the amplifier in its two operating modes. The tests proved that the embedded amplifier can work both in Charge Sensitive Mode with an estimated gain equal to 108±2 mV/fC and a linear response up to ≃ 25 fC and in Transimpedance mode with a conversion factor of 2.10±0.01mVmGy/s
Comparative analysis of diamond graphitization approaches for 3D electrode fabrication
No full textThis work presents a comprehensive comparative study of different methods to induce graphitization in diamond, with a focus on their impact on the material properties and potential applications. The employed graphitization methods vary between laser writing-based strategies using Bessel beams and irradiation-based methods incorporating keV broad beam implantation, focused Gallium ion beam lithography, and a thermal-based graphitization method. The study emphasizes the effects of annealing time in the thermal method and the influence of the induced graphite on the material electrical properties for all the employed methods. A key highlight of this work is the exploration of easy and innovative approaches to create 3D conductive patterns within diamonds. This is achieved through combining laser-based and ion beam methods, as well as optimizing ion beam parameters to enable the creation of 3D conductive patterns. Each method presents unique advantages: thermal annealing enables simple surface graphitization, ion implantation allows precise depth control for sub-surface structures, and Bessel beam laser writing facilitates direct 3D in-bulk modification without mechanical movement or masking. This comparative insight helps to identify the most suitable technique depending on the required electrode geometry and device integration constraints. The outcomes of each graphitization approach showcase unique benefits and limitations, giving various possibilities to employ diamonds in electrical applications. By building on prior research, this study offers an extended perspective on integrating advanced fabrication techniques to expand the use of diamonds in emerging technologies
An LLM-as-a-judge Approach for Scalable Gender-Neutral Translation Evaluation
Get PDFGender-neutral translation (GNT) aims to avoid expressing the gender of human referents when the source text lacks explicit cues about the gender of those referents. Evaluating GNT automatically is particularly challenging, with current solutions being limited to monolingual classifiers. Such solutions are not ideal because they do not factor in the source sentence and require dedicated data and fine-tuning to scale to new languages. In this work, we address such limitations by investigating the use of large language models (LLMs) as evaluators of GNT. Specifically, we explore two prompting approaches: one in which LLMs generate sentence-level assessments only, and another, akin to a chain-of-thought approach, where they first produce detailed phrase-level annotations before a sentence-level judgment. Through extensive experiments on multiple languages with five models, both open and proprietary, we show that LLMs can serve as evaluators of GNT. Moreover, we find that prompting for phrase-level annotations before sentence-level assessments consistently improves the accuracy of all models, providing a better and more scalable alternative to current solutions
Aspect Ratio‐Engineered Ru‐Integrated W18O49: Controlled Growth and Enhanced Electrocatalytic Activity
Get PDFThe sustainable production of renewable fuels and feedstocks is currently constrained by the slow kinetics of anodic oxygen evolution reaction (OER). Precious metal-based catalysts such as Ir suffer from stability issues as well as high capital cost. To enforce the future of green hydrogen production, this study develops Ru-integrated W18O49 nanowires (NWs), as an efficient and stable OER electrocatalyst. This study obtains Ru-W18O49 NWs by a combined physical vapor deposition–chemical vapor deposition approach. It discovers the NWs growth mechanism, characterized by two different growth kinetics. Herein, it finds that the integration of just 3% of Ru in the oxygen-deficient W18O49 NWs remarkably increases the number of active catalytic sites during OER, showing faster kinetics (60 mV dec−1) and a reduced overpotential of 360 mV at 10 mA cm−2. The electrode's observed catalytic performance and long-term durability over 36 h (12 h each at 10, 30, and 100 mA cm−2) combined with the versatility of the two-step synthetic route, are a promising research approach for future industrial applications
A Virtual Reality-Based Telerehabilitation System for Reminiscence Therapy in Individuals with Cognitive Impairments.
No full textSearch for light neutral particles decaying promptly into collimated pairs of electrons or muons in pp collisions at √s = 13 TeV with the ATLAS detector
No full textA search for a dark photon, a new light neutral particle, which decays promptly into collimated pairs of electrons or muons is presented. The search targets dark photons resulting from the exotic decay of the Standard Model Higgs boson, assuming its production via the dominant gluon-gluon fusion mode. The analysis is based on
of data collected with the ATLAS detector at the Large Hadron Collider from proton-proton collisions at a center-of-mass energy of 13
. Events with collimated pairs of electrons or muons are analysed and background contributions are estimated using data-driven techniques. No significant excess in the data above the Standard Model background is observed. Upper limits are set at 95% confidence level on the branching ratio of the Higgs boson decay into dark photons between 0.001% and 5%, depending on the assumed dark photon mass and signal model