HAL-Polytechnique
Not a member yet
51406 research outputs found
Sort by
Type I and type II superconductivity in a quasi-2D Dirac metal
No full textInternational audienceDirac material LaCuSb 2 shows anisotropic superconducting response to applied magnetic fields.Abstract: We explore bulk superconducting phase in single crystals of the Dirac material LaCuSb2 prepared by the self-flux method. Magnetization, muon spin relaxation measurements, and density functional theory, show the Dirac nodal line Fermi surfaces give rise to type-II superconductivity for magnetic fields applied along the a-axis, and type-I superconductivity for fields along the c-axis. Both chemical and hydrostatic pressure drastically suppress the superconducting transition. We find multiband superconductivity evidenced by a precipitous drop in the electronic specific heat capacity and high-pressure susceptibility for T* < Tc/3. Our work demonstrates dirty-limit, weak-coupling multiband superconductivity in LaCuSb2, and highlights the role of Dirac fermions on its anisotropic character
Mathematical model linking telomeres to senescence in Saccharomyces cerevisiae reveals cell lineage versus population dynamics
No full textInternational audienceTelomere shortening ultimately causes replicative senescence. However, identifying the mechanisms driving replicative senescence in cell populations is challenging due to the heterogeneity of telomere lengths and the asynchrony of senescence onset. Here, we present a mathematical model of telomere shortening and replicative senescence in Saccharomyces cerevisiae which is quantitatively calibrated and validated using data of telomerase-deficient single cells. Simulations of yeast populations, where cells with varying proliferation capacities compete against each other, show that the distribution of telomere lengths of the initial population shapes population growth, especially through the distribution of cells’ shortest telomere lengths. We also quantified how factors influencing cell viability independently of telomeres can impact senescence rates. Overall, we demonstrate a temporal evolution in the composition of senescent cell populations—from a state directly linked to critically short telomeres to a state where senescence onset becomes stochastic. This population structure may promote genome instability and facilitate senescence escape
Modélisation actuarielle du risque cyber en assurance : fréquence, contagion et accumulation systémique via les processus de Hawkes avec excitation externe
No full textThis thesis examines the actuarial modeling of cyber risk, a fast-growing systemic risk driven by contagion and accumulation phenomena. These features challenge traditional approaches based on the assumption of independence between policyholders.The thesis is organized into six chapters. The first two give a general introduction to the subject, presenting the main issues and methodological contributions.The third chapter introduces a Hawkes process with external excitation, incorporating the impact of software vulnerabilities and random marks to capture attacks heterogeneity. Several calibration methods are developed and compared.The fourth chapter applies this model to real data combining cyber incidents (Hackmageddon) and software vulnerabilities (NVD). It shows that ignoring vulnerabilities leads to an overestimation of endogeneity, and proposes a two-phase extension to reflect protection measures and simulate cyber-pandemic scenarios.The fifth chapter examines heterogeneity in attack propagation. An adaptation of the CART algorithm to Hawkes process trajectories is proposed, enabling the classification of attack dynamics and providing insights for insurance portfolios structuring.The sixth chapter addresses cyber stress tests in insurance. We propose closed-form formulas to assess the average impact of stress-test scenarios on the total portfolio loss.Overall, this thesis proposes new methodologies advances for improved quantification of cyber risk and introduces new tools to support decision-making in insurance.Cette thèse porte sur la modélisation actuarielle du risque cyber, un risque systémique en forte croissance marqué par des phénomènes de contagion. Ces spécificités remettent en cause les approches classiques fondées sur l'indépendance entre assurés.Cette thèse est structurée en six chapitres. Les deux premiers introduisent le sujet, présentent les problématiques et les contributions méthodologiques.Le troisième chapitre propose un processus de Hawkes avec excitation externe intégrant l'impact des vulnérabilités informatiques et des marques aléatoires pour représenter l'hétérogénéité des attaques. Plusieurs méthodes de calibrage sont développées et comparées.Le quatrième chapitre applique ce modèle à des données réelles (Hackmageddon, NVD), montrant que négliger les vulnérabilités conduit à surestimer l'endogénéité, et propose une extension en deux phases pour représenter l'effet des mesures de protection et simuler des scénarios de cyber-pandémie.Le cinquième chapitre aborde l'hétérogénéité de propagation des attaques. Une adaptation de l'algorithme CART aux trajectoires de Hawkes permet de classer les attaques et d'éclairer la structuration des portefeuilles d'assurance.Le sixième chapitre aborde les stress tests cyber en assurance et propose des formules fermées pour évaluer l'impact de ces scénarios sur la perte totale.Dans son ensemble, cette thèse propose des avancées méthodologiques pour mieux quantifier le risque cyber et de nouveaux outils pour la décision assurantielle
Collagen-based biomimetic tubular materials for vascular and tracheal applications
No full textInternational audienceBiomimetic scaffolds have many reasons to be chosen over synthetic grafts as they show low biological response, which means less immune response to the implanted material, and can promote regeneration. Type I collagen is one of the main components in the extracellular matrix and therefore a protein of choice to work with in the tissue engineering field. Our approach combining ice-templating method and topotactic fibrillogenesis (1) enabled to obtain collagen-based double layered materials. The distinct structures reproduce the multi-scale architecture of native tissues without the use of any chemical crosslinker, and reach mechanical and water-tightness properties relevant for in vivo implantation. Tubular grafts being a challenge to fabricate, there is actually a high clinical demand for vascular implants of limited diameter and for grafts for tracheal repair. For instance, the most promising solution to repair the trachea today is replacement by cryopreserved aorta, as demonstrated by Martinod et al. (2,3). Previous work was done by Martinier et al. (4) on vascular size scaffold simplifying the native three-layered vascular structure to a double layer (DL) model. This process was then adapted for a vascular model (Figure 1A) and then on a larger scale, for tracheal applications (Figure 1B). The scaffolds prepared with this method were characterized with mechanical tests, confocal and TEM imaging, were seeded by HUVECs and hMSCs and tested by surgeons.Type I collagen at 40 mg/mL was poured into cylindrical PTFE mold of 8 mm diameter for vascular size and a 20 mm one for tracheal size. A metallic tube of 8 mm, 17 mm for tracheal size, was then added inside. Sample was then frozen in liquid nitrogen and was placed into an NH3 vapor chamber to fibrillate collagen. Acidic collagen at 40 mg/mL was added inside the tube and a PTFE rod of 4 mm for vascular caliber, 15 mm for the trachea, was inserted. Sample was put in PBS5X for stabilization then matrix was collected. Tubular uniaxial tensile testing and inflation experiments were performed on a custom-made platform, with piglet carotid arteries used as controls. For biological response, HUVECs were seeded on luminal side of the tube and cadherin expression and shape of the cells were observed using confocal and hMSCs were seeded on the external surface.The confocal image of trachea material show that both layers are homogenously sealed all the way in the transversal section, with a porous homogenous layer on the external while the inner layer is smooth (Figure 1C). The TEM micrographs show dense collagen material with the typical striation of fibrillated collagen found in the extracellular matrix (Figure 1D). Confocal image of HUVECs display expression of cadherin implying junction formation between cells, which corresponds to the formation of an epithelium on the lumen (Figure 1E), while hMSCs penetrate the scaffold through the porosity of the matrix (Figure 1F). Longitudinal elongation measurements show that native artery’s Young’s modulus is around 80 kPa whereas our material’s is around 30 kPa (Figure 2A,B). Compliance, which values the radial deformation under pressure differences between diastole and systole, has been measured in hypotensive, normotensive and hypertensive regimes (Figure 2C). The compliance at 80-120 mmHg measured for the native artery is around 11% whereas for the scaffold the value is around 6%. Under inflation, scaffolds did not leak nor break until more than 200 mmHg.Smooth and porous layers are homogeneously adhesive to each other all the way through the sample. Fibrils are present in the scaffold, which means sample is highly concentrated and structured at the molecular level. Seeded cells have sufficient access to nutrients to colonize the substrate on the porous side and form an endothelium on the luminal side. Having a higher Young’s modulus, native artery is more resistant to the stretching compared to the non seeded vascular scaffold, which mechanical properties are only due to processing of collagen in non denaturing conditions, in the absence of cells. The double layered material has a Young’s modulus of 30 kPa which correlates with the values for the porous and non porous material, 20 kPa and 67 kPa respectively (4), which indicates that the double layer structure enables to combine the properties of each layer. Compliance is in the same range for the native artery and the double layer scaffold, which is of uttermost importance for vascular applications. For tracheal applications, the surgeons strategy is to use the scaffold with a stent during cell colonization and while cartilage forms, which would compensate these mechanical differences as in the actual procedures.The obtained tubular biomaterials, which display appropriate mechanical properties and promote cell colonization and endothelium formation, open an exciting pathway for obtaining off-the-shelf grafts for vascular and tracheal applications.(1) Rieu C et al, 2019, 10.1021/acsami.9b03219(2) Martinod E et al. 2010, 10.1016/j.rmr.2010.04.001(3) Martinod E et al. 2018, 10.1001/jama.2018.465(4) Martinier I et al. 2024, 10.1039/D3BM01808
Signalling for electricity demand response: When is truth telling optimal?
No full textUtilities and transmission system operators (TSO) around the world implement demand response programs for reducing electricity consumption by sending information on the state of balance between supply demand to end-use consumers. We construct a Bayesian persuasion model to analyse such demand response programs. Using a simple model consisting of two time steps for contract signing and invoking, we analyse the relation between the pricing of electricity and the incentives of the TSO to garble information about the true state of the generation. We show that if the electricity is priced at its marginal cost of production, the TSO has no incentive to lie and always tells the truth. On the other hand, we provide conditions where overpricing of electricity leads the TSO to provide no information to the consumer
Design to production of HGCROC3 and H2GCROC3: radiation-hard front-end ASICs for the CMS HGCAL
No full textInternational audienceThe CMS High Granularity Calorimeter (HGCAL), developed for the High luminosity phase of the LHC (HL-LHC), uses custom ASICs — HGCROC3 and H2GCROC3 — to read out silicon sensors and SiPM-on-tile modules. These chips provide precise charge and timing measurements, digital processing for triggering, and are designed to operate in harsh radiation conditions. Version 3 of the chips implements all final features, with sub-versions a–d addressing bugs and improving radiation tolerance. Extensive testing has been performed in lab and beam environments. The proceedings covers chip design, performance, and SEE-related issues and fixes
FIT-IRSA: Feedback-Integrated Two-Phase IRSA with Deep Reinforcement Learning
No full textInternational audienceEfficient random access can be used in scenarios with a massive number of IoT devices. Among modern random access protocols, Irregular Repetition Slotted ALOHA (IRSA) offers excellent asymptotic performance (for large frame sizes), but its finite-frame efficiency is lower and difficult to optimize analytically. In this work, we introduce limited mid-frame feedback to better coordinate users and improve performance: Feedback-Integrated Two-phase IRSA (FIT-IRSA). We formulate IRSA with feedback as a deep reinforcement learning (DRL) problem. Using policy gradient methods, we learn transmission strategies that improve throughput under varying loads, as demonstrated in our simulation results. This provides a practical alternative to classical density-evolution-based optimization, which applies mainly to large frames
Algorithm- and Data-Dependent Generalization Bounds for Score-Based Generative Models
No full textInternational audienceScore-based generative models (SGMs) have emerged as one of the most popular classes of generative models. A substantial body of work now exists on the analysis of SGMs, focusing either on discretization aspects or on their statistical performance. In the latter case, bounds have been derived, under various metrics, between the true data distribution and the distribution induced by the SGM, often demonstrating polynomial convergence rates with respect to the number of training samples. However, these approaches adopt a largely approximation theory viewpoint, which tends to be overly pessimistic and relatively coarse. In particular, they fail to fully explain the empirical success of SGMs or capture the role of the optimization algorithm used in practice to train the score network. To support this observation, we first present simple experiments illustrating the concrete impact of optimization hyperparameters on the generalization ability of the generated distribution. Then, this paper aims to bridge this theoretical gap by providing the first algorithmic- and data-dependent generalization analysis for SGMs. In particular, we establish bounds that explicitly account for the optimization dynamics of the learning algorithm, offering new insights into the generalization behavior of SGMs. Our theoretical findings are supported by empirical results on several datasets
4f-orbital covalency enables a single-crystal-to-single-crystal ring-opening isomerization in a CeIV–cyclopropenyl complex
No full textInternational audienceMetal–ligand bonding interactions for f-element compounds are typically highly polarized with only minor covalent character. Whereas the 5d/6d orbitals are known to be chemically accessible for dative bonding, recent quantum chemical and spectroscopic analyses have indicated appreciable 4f/5f-orbital involvement in certain metal–ligand bonds. However, 4f-orbital covalency has not been compellingly linked to distinctive modes of chemical reactivity via rigorous comparative study and mechanistic investigation. Here a series of MIV–cyclopropenyl complexes (M = Ti, Zr, Ce, Hf, Th) are described, wherein the cerium congener exhibits a 4f-covalent Ce=Cα interaction, causing a ring-opening isomerization reaction through a single-crystal-to-single-crystal transformation. The results provide evidence for 4f-orbital covalency by demonstrating its expression in the reactivity of an f-element complex within an isostructural series of tetravalent d- and f-block metal complexes. They also provide new directions for the study of orbital covalency effects of molecular compounds in solid-state chemical transformations
The Cost of Skeletal Call-By-Need, Smoothly
No full textInternational audienceSkeletal call-by-need is an optimization of call-by-need evaluation also known as "fully lazy sharing": when the duplication of a value has to take place, it is first split into "skeleton", which is then duplicated, and "flesh" which is instead kept shared. Here, we provide two cost analyses of skeletal call-by-need. Firstly, we provide a family of terms showing that skeletal call-by-need can be asymptotically exponentially faster than call-by-need in both time and space; it is the first such evidence, to our knowledge. Secondly, we prove that skeletal call-by-need can be implemented efficiently, that is, with bi-linear overhead. This result is obtained by providing a new smooth presentation of ideas by Shivers and Wand for the reconstruction of skeletons, which is then smoothly plugged into the study of an abstract machine following the distillation technique by Accattoli et al