HAL Université de Toulouse, et Toulouse INP
Not a member yet
159308 research outputs found
Sort by
First observation of quantum oscillations by transport measurements in semi-destructive pulsed magnetic fields up to 125 T
No full textInternational audienceHigh magnetic fields have proven instrumental in exploring the physical properties of condensed matter, leading to groundbreaking discoveries, such as the quantum Hall effect in 2D heterostructures and quantum oscillations in cuprate superconductors. The ability to conduct precise measurements at progressively higher magnetic fields continues to push the frontiers of knowledge and enable new discoveries. In this work, we present the development of a microwave technique for performing two-point transport measurements in semi-destructive pulsed magnetic fields (up to 125 T) and at low temperatures (down to 1.5 K) with unprecedented sensitivity. This new setup was tested on a variety of samples. We present results on the metal–insulator transition in InAs and we report notably the first observation of Shubnikov–de Haas oscillations in WTe2 at magnetic fields beyond 100 T
A Semantic Framework for the Contextual Interpretation of ADHD Symptom Manifestations
No full textInternational audienceAttention Deficit Hyperactivity Disorder is a neurodevelopmental disorder whose manifestations vary significantly depending on the context. This situational variability poses major challenges for assessing and understanding symptoms, particularly outside clinical environments. In this work, we propose a framework that integrates a contextual vision to enrich medical information. The framework is composed of three main components: a modular ontology that formalizes both medical and contextual dimensions of ADHD; a multi-agent system powered by large language models for automatically extracting and populating knowledge from heterogeneous data sources; and a clinical rule-based reasoning mechanism capable of inferring high-level interpretations from instantiated data. Experimental results demonstrate the framework's ability to generate accurate, context-sensitive interpretations of symptom manifestations. This approach lays the groundwork for more personalized, explainable, and context-aware patient monitoring, with promising applications in intelligent healthcare systems
Spermidine reproduces the anti‐inflammatory effects of intermittent fasting and prevents urate and calcium pyrophosphate crystal‐induced inflammation
No full textInternational audienceBackground Gout due to the formation of monosodium urate (MSU) crystals and calcium pyrophosphate (CPP) deposition disease are two major types of microcrystalline pathologies in adults. They are responsible for recurrent flares that rely on interleukin (IL)‐1β via activation of the NLRP3 inflammasome. Intermittent fasting (IF) is a non‐pharmacological intervention that improves age‐related diseases and reduces inflammation. Methods In air pouch model, crystal‐induced inflammation was compared between mice fed ad libitum and mice under IF. Systemic (liver and serum) and local (air pouch cavity) modifications were evaluated by metabolomics analysis. The anti‐inflammatory potential of metabolites was tested in vitro and in vivo. Results we observe that 2 non‐consecutive days of fasting during one week significantly prevents the inflammation provoked by MSU and CPP crystal injection into air pouch cavity in mouse model. This short‐term fasting is associated with increased serum abundances of numerous anti‐inflammatory metabolites, including β‐hydroxybutyrate and spermidine (SPD), a polyamine able to reproduce biological effects of IF. Conversely, crystal stimulation in ad libitum fed mice decreases the local production of these metabolites in the air pouch membrane. Supplementation of SPD reproduces the anti‐inflammatory effect of IF and prevents crystal‐induced inflammation through inhibition of NF‐κB and NLRP3 inflammasome. Finally, downregulation of SAT1 (spermidine/spermine acetyltransferase 1), which encodes the enzyme that degrades SPD, reduces IL‐1β production induced by crystal stimulation. Conclusion In summary, we have identified several metabolites that recapitulate the anti‐inflammatory effects of IF and could be used as IF mimetics to prevent microcrystal inflammation
Evaluation of a Chain of Decoupling Capacitors for Fast Switching GaN HEMT Transistors
No full textInternational audienceDC voltage decoupling across a switching halfbridge has attracted significant attention, particularly with the adoption of GaN HEMT transistors, whose high switching speed exacerbates voltage overshoot and ringing. A design approach based on a cascade (chain) of decoupling capacitors is proposed to address the latter challenge. Counter-intuitively, DC voltage decoupling improves when specific capacitors are associated in a succession of filters. This paper presents the theoretical analysis of the proposed configuration, followed by experimental validation using a double-pulse testbench. The results confirm the positive impact on transistor switching performance of the decoupling capacitor arrangement
Search for signatures of electroweakinos with photons, jets, and large missing transverse momentum in TeV pp collisions with the ATLAS detector
No full textInternational audienceA search for final states characterised by at least one isolated high transverse-momentum photon, jets and large missing transverse momentum is presented. Such a final state might occur in gauge-mediated supersymmetric models where a pair of binos and higgsinos mix to form neutralinos, one of which decays into a photon plus a gravitino while the other decays into a Higgs boson, a Z boson or a photon, plus a gravitino. The search is performed using the full Run-2 data sample of 140 fb of TeV proton-proton collisions collected by the ATLAS detector at the Large Hadron Collider. No significant excess of events is observed above the Standard Model prediction and model-dependent exclusion limits at the 95% confidence level are set. These limits are interpreted in terms of the masses of gauginos, which are excluded up to 1.2 TeV depending on their branching ratios, and on their branching ratios as a function of their mass
Hybridization in van der Waals Epitaxy of PtSe 2 /h-BN and PtSe 2 /Graphene Heterostructures
No full textInternational audiencevan der Waals (vdW) heterostructures, which combine bidimensional materials of different properties, enable a range of quantum phenomena. Herein, we present a comparative study between the electronic properties of mono- and bilayer of platinum diselenide (PtSe2) grown on hexagonal boron nitride (h-BN) and graphene substrates using molecular beam epitaxy (MBE). Using angle-resolved photoemission spectroscopy (ARPES) and density functional theory (DFT), the electronic structure of PtSe2/graphene and PtSe2/h-BN vdW heterostructures is investigated in a systematic manner. In contrast to PtSe2/h-BN, the electronic structure of PtSe2/graphene reveals the presence of interlayer hybridization between PtSe2 and graphene, which is evidenced by minigap openings in the π-band of graphene. Furthermore, our measurements show that the valence band maximum (VBM) of monolayer PtSe2 is located at the Γ point with different binding energies of about −0.9 and −0.55 eV relative to the Fermi level on h-BN and graphene and substrates, respectively. Our results represent a significant advance in the understanding of electronic hybridization between TMDs and different substrates, and they reaffirm the crucial role of the substrate in any nanoelectronic applications based on van der Waals heterostructures
Carbon chain diversity in L1544 and IRAS 16293–2422: an astrochemical pathfinder study for the SKAO
No full textInternational audienceAstrochemical observations have revealed a surprisingly high level of chemical complexity, including long carbon chains, in the earliest stages of Sun-like star formation. The origin of these species and whether they undergo further growth, possibly contributing to the molecular complexity of planetary systems, remain open questions. We present recent observations performed using the 100-m Green Bank Telescope of the prestellar core L1544, and the protostellar system IRAS 16293–2422. In L1544, we detected several complex carbon-bearing species, including CS, CS, CN, c-CH, CH, and CH, complementing previously reported emission of cyanopolyynes. In IRAS 16293–2422, we detected c-CH and, for the first time, HCN. Thanks to the high spectral resolution, we refine the rest frequencies of several c-CH and CH transitions. We perform radiative transfer analysis, highlighting a chemical difference between the two sources: IRAS 16293–2422 shows column densities 10 to 100 times lower than L1544. We perform astrochemical modelling, employing an up-to-date chemical network with revised reaction rates. Models reproduce the general trends, with cyanopolyyne and polyynyl radical abundances decreasing as molecular size increases, but underestimate the abundances of cyanopolyynes longer than HCN by up to two orders of magnitude. Current models, which include the dominant neutral–neutral formation routes, cannot account for this discrepancy, suggesting that the chemical network is incomplete. We propose that additional ion–molecule reactions are crucial for the formation of these species. Developing a more comprehensive chemical network for long carbon chains is essential for accurately interpreting present and future observations
Développement d’une application à destination des propriétaires canins aidant au suivi des besoins du chien
No full textThis thesis is about creating a mobile app for dog owners to make it easier to find and keep track of important information about their dog’s health, food, exercise, and training. The app is meant to support the advice given during vet visits, with tips that match the dog’s lifestyle and needs. It’s divided into several sections: a health tracker with reminders for vaccines and parasite treatments; food tips that cover weight, poisoning risks, and dental care; fun games and activity ideas based on the dog’s age and breed; and info on how dogs communicate and how to help them socialize. The goal is to help owners better understand what their dogs need and to give general health advice without replacing a real vet’s opinion.Cette thèse porte sur la conception d’une application mobile destinée aux propriétaires de chiens, visant à centraliser et faciliter l’accès aux informations essentielles concernant la santé, l’alimentation, l’activité et l’éducation de leur animal. L’application a pour but de compléter les conseils donnés lors des consultations vétérinaires, adaptés en fonction de l’animal et de son mode de vie. Elle se structure en plusieurs modules : suivi de santé avec rappels de vaccination et de traitements antiparasitaires, conseils alimentaires incluant le poids, les risques d’intoxications et l’hygiène bucco-dentaire, idées de jeux et d’activités adaptées à l’âge et à la race du chien, informations sur la communication canine et la sociabilisation. L’enjeu est de sensibiliser et d’éduquer les propriétaires sur les besoins de leurs animaux, ainsi que de prodiguer des conseils portant sur la médecine préventive au sens large sans se substituer à un avis professionnel
Sequential Conformal Risk Control for Safe Railway Signaling Detection
No full textInternational audienceAs machine learning becomes a more common tool in industry, its needs for certification increase. Conformal Prediction (Vovk et al., 2005), a framework for construction of prediction sets with tight coverage guarantees at any desired error rate, is an ideal tool for this purpose. However, adapting conformal methods to complex computer vision pipelines and providing appropriate guarantees is still a challenging task. Indeed, conformal approaches to object detection are often restricted to subtasks: often localization, and sometimes classification. In this study, we apply the comprehensive framework from Andéol et al. (2025) to the safety-critical task of railway signaling detection
Residual Prophet Inequalities
No full textWe introduce a variant of the classic prophet inequality, called \emph{residual prophet inequality} (RPI). In the RPI problem, we consider a finite sequence of nonnegative independent random values with known distributions, and a known integer . Before the gambler observes the sequence, the top values are removed, whereas the remaining values are streamed sequentially to the gambler. For example, one can assume that the top values have already been allocated to a higher-priority agent. Upon observing a value, the gambler must decide irrevocably whether to accept or reject it, without the possibility of revisiting past values. We study two variants of RPI, according to whether the gambler learns online of the identity of the variable that he sees (FI model) or not (NI model). Our main result is a randomized algorithm in the FI model with \emph{competitive ratio} of at least , which we show is tight. Our algorithm is data-driven and requires access only to the largest values of a single sample from the input distributions. In the NI model, we provide a similar algorithm that guarantees a competitive ratio of . We further analyze independent and identically distributed instances when . We build a single-threshold algorithm with a competitive ratio of at least 0.4901, and show that no single-threshold strategy can get a competitive ratio greater than 0.5464