Archive ouverte de Centrale Lyon
Not a member yet
32420 research outputs found
Sort by
Villain action in lattice gauge theory
No full textInternational audienceWe prove that Villain interaction applied to lattice gauge theory can be obtained as the limit of both Wilson and Manton interactions on a larger graph which we call the {\em carpet graph.} This is the lattice gauge theory analog of a well-known property for spin models where Villain type interactions are the limit of spin systems defined on a {\em cable graph}. Perhaps surprisingly in the setting of lattice gauge theory, our proof also applies to non-Abelian lattice theory such as -lattice gauge theory and its limiting Villain interaction. In the particular case of an Abelian lattice gauge theory, this allows us to extend the validity of Ginibre inequality to the case of the Villain interaction
Faithful Interpretation of Protein Structures through Weighted Persistent Homology Improves Evolutionary Distance Estimation
No full textInternational audienceAbstract Phylogenetic inference is mainly based on sequence analysis and requires reliable alignments. This can be challenging, especially when sequences are highly divergent. In this context, the use of three-dimensional protein structures is a promising alternative. In a recent study, we introduced an original topological data analysis method based on persistent homology to estimate the evolutionary distances from structures. The method was successfully tested on 518 protein families representing 22,940 predicted structures. However, as anticipated, the reliability of the estimated evolutionary distances was impacted by the quality of the predicted structures and the presence of indels in the proteins. This paper introduces a new topological descriptor, called bio-topological marker (BTM), which provides a more faithful description of the structures, a topological analysis for estimating evolutionary distances from BTMs, and a new weight-filtering method adapted to protein structures. These new developments significantly improve the estimation of evolutionary distances and phylogenies inferred from structures
Quantum Hall Edges Beyond the Plasma Analogy
No full textInternational audienceWe demonstrate that the plasma analogy is generally unreliable at predicting the edge properties of quantum Hall (QH) states, as it fails to account for the local edge velocity. This discrepancy arises from a fundamental difference between QH droplets and Coulomb gases (CGs): the former are incompressible liquids subject to area-preserving deformations, while the latter are governed by electrostatics and thus involve conformal transformations. We illustrate this QH/CG mismatch across various examples and show its impact on physical quantities, measurable in both solid-state samples and quantum simulators. Specifically, we discuss fluctuations of local observables and their connection to state-of-the-art microwave absorption experiments
Slope Compensation and Bifurcation in a DC-DC, Single-Input, Multiple-Output, CMOS Integrated Converter Under Current-Mode and Comparator-Based Hybrid Control
No full textInternational audienceSingle-Input, Multi-Output (SIMO) converters present significant challenges when operated under current-mode control, due to their strongly non-linear dynamics and susceptibility to bifurcation phenomena. To mitigate the effects on the converter’s steady-state, a double slope compensation solution is proposed. The compensation parameters play a critical role in shaping the system dynamics and rejecting the susceptibility to bifurcation. This paper proposes a detailed analysis methodology to investigate the design parameter space regarding the slope compensations with respect to bifurcation phenomena. The approach is validated on a CMOS integrated converter, where theoretical predictions are compared to the simulation results of a full transistor-level model of the circuit
Generative AI for Low-Resource NLP: Evaluating Khmer Data Generation for Homophone Spelling Correction
No full textInternational audienceThis study demonstrates that generative AI can transform data scarcity from a persistent obstacle into a powerful opportunity for advancing NLP in low-resource languages, with Khmer as the primary focus. The shortage of high-quality, domain-specific datasets, compounded by Khmer's complex syntax and abundance of homophones, has long limited the development of effective language technologies. To address this, we evaluated five state-of-the-art generative AI models: Qwen3-235B-A22B, ChatGPT's GPT-4o, Claude Sonnet 4, DeepSeek-V3 (R1), and Gemini 2.5 Pro, on their ability to produce high-quality Khmer text for training Sor-Ser, a homophone spelling correction system. Our evaluation measured accuracy, diversity, grammatical precision, semantic relevance, and redundancy. GPT-4o consistently emerged as the top performer, generating diverse, syntactically precise, and contextually accurate sentences with minimal redundancy, making it the most suitable for this task. Claude Sonnet 4 and Qwen3-235B-A22B also showed strengths in specific dimensions, suggesting their potential in targeted applications. These results provide strong evidence that generative AI can produce scalable, high-quality datasets tailored to linguistic and domain-specific needs. By turning data scarcity into opportunity, this work lays the groundwork for bridging the gap between high-and low-resource languages, driving more inclusive, equitable, and sustainable NLP development
Hybrid Si/III-V quantum cascade lasers integrated on a phase-matched mid-infrared silicon photonic platform
No full textInternational audienceThe increasing demand for mid-infrared (MIR) photonic integrated devices in spectroscopic applications has driven the development of essential building blocks for chemical sensing, including quantum cascade lasers (QCLs), MIR silicon photonic platforms and integrated detectors. A fully integrated MIR silicon photonic chip would enable cost-effective mass production using CMOS-compatible fabrication, paving the way for consumer-grade MIR devices and large-scale onsite sensing applications. However, the integration of QCLs with MIR silicon platforms remains a major challenge due to inefficient optical coupling between the III-V active region and silicon waveguides, leading to high insertion losses and reduced device performance. In this work, we demonstrate the heterogeneous integration of a QCL onto a believed to be novel high-index-contrast, phase-matched silicon on nitride on insulator (SONOI) photonic platform using molecular bonding. By leveraging a phase-matching condition, we implement an adiabatic coupling scheme that ensures efficient optical power transfer from the III-V active region to the silicon waveguides, overcoming a key limitation of previous approaches. The resulting hybrid distributed-feedback (DFB) QCL exhibits single-mode emission at 4.315 µm and operates in pulsed mode up to 72 °C. This advancement opens new possibilities for fully integrated MIR photonic circuits, with potential applications in environmental monitoring, biomedical diagnostics, and industrial sensing
Non-stable -functors of discrete valuation rings containing a field
No full textLet k be a field, and let G be a simply connected semisimple k-group which is isotropic and contains a strictly proper parabolic k-subgroup P . Let D be a discrete valuation ring which is a local ring of a smooth algebraic curve over k. Let K be the fraction field of D. We show that the corresponding non-stable K 1 -functor(for G and P,, also called the Whitehead group of G) coincide over D and K.. As a consequence, K_G^P 1 (D) coincides with the (generalized) Manin's R-equivalence class group of G(D)
Commande et optimisation d'un convertisseur DAB embarqué : analogie avec un moteur synchrone
No full textInternational audienc
Towards a wide bandgap absorber: structural, morphological, and optical investigation of Ag-alloyed Cu 2 ZnSnS 4 thin films
No full textInternational audienceControlled synthesis of Ag-alloyed Cu 2 ZnSnS 4 (ACZTS) thin films with wide range of Ag/Cu ratios via a solution process enables phase transition, lattice expansion, and bandgap tuning, yielding stable absorbers for next-generation photovoltaics
Effects of shear stress on mesenchymal stem cells of patients with osteogenesis imperfecta
No full textInternational audienceIntroductionOsteogenesis imperfecta (OI) is a rare genetic bone disorder, mainly caused by autosomal dominant mutations of the COL1A1 or COL1A2 genes that encode the alpha chains of type 1 collagen. In severe forms and in nonambulatory patients, for whom physical exercise is difficult, exposing the bone to mechanical stimuli by promoting movement, especially with physiotherapy and mobility aids, is an essential part of clinical practice. However, the effects of mechanical stimulation at the cellular level remain unknown for this disease.HypothesisThe study hypothesis was that human mesenchymal stem cells (hMSCs) from patients with OI were as sensitive to mechanical stimulation as those from healthy patients, validating the current clinical practice.Materials and methodshMSCs were harvested from 3 healthy control subjects and 3 patients with OI during an elective osteotomy of a long bone of the lower limb. The healthy and OI hMSCs were then exposed to mechanical stimuli, such as intermittent shear stress of 0, 0.7, 1.5, and 3 Pascal (Pa) at a frequency of 2.8 Hertz (Hz) for 30 min using a commercial ibidi system. The immediate early gene expression of themechanosensitive prostaglandin-endoperoxide synthase 2 (PTGS2) was examined 1 h after stimulation to determine the best level of mechanical stimulation. The expression of 7 other mechanosensitive genes was also examined for this level of mechanical stimulation after applying intermittent shear stress at 1.5 Pa.ResultsIn all hMSCs, mechanical stimulation induced PTGS2 gene overexpression with a maximum after exposure to intermittent shear stress of 1.5 Pa and without significant differences between OI and healthy donors. Except for fibroblast growth factor 2, gene expression in OI donors was found to be significantly different from that in hMSCs not exposed to shear stress. Moreover, the relative expression associated with mechanical stimulation was not significantly different between healthy and OI donors for most other genes.DiscussionThis is the first study to demonstrate that hMSCs from patients with OI are as sensitive to mechanical shear stress as those from healthy donors. The mechanical stress that resulted in the greatest change in the expression of PTGS2 in patients with OI was similar to that previously reported in the literature for healthy subjects. These findings are an important step toward further fundamental research aimed at confirming the effects of mechanical stress at the cellular level over the long term and, more importantly, toward developing clinical protocols for delivering mechanical stimuli to these patients