324139 research outputs found
Sort by
QAOA-Driven PMU placement optimization with graph learning-based parameter initialization refinement
No full textWith the significant expansion of renewable energy integration, the scale of the power grid also increases rapidly. To effectively monitor the operational state of large-scale power grids, optimizing the placement of Phasor Measurement Units (PMUs) is a critical research focus. Optimizing PMU Placement (OPP) is a typical combinatorial optimization problem with NP-hard complexity, which brings substantial challenges to classical computers. The sheer scale of modern grids forces classical solvers into prohibitive runtimes and sub-optimal local minima, degrading both execution speed and solution quality. Recent advances in quantum computing have opened new opportunities for tackling combinatorial optimization problems, particularly through the Quantum Approximate Optimization Algorithm (QAOA). However, QAOA operates as a hybrid quantum–classical framework, where determining the optimal parameters depends on a classical optimization process that remains computationally challenging and inherently NP-hard. On the other hand, in the Noisy Intermediate-Scale Quantum (NISQ) era, obtaining optimal optimization results typically requires a large number of quantum measurement shots. In this work, we propose a graph-learning-based strategy to provide QAOA with guided parameter initialization, enabling effective operation under limited quantum resources, particularly when the number of available measurement shots is restricted. Both the OPP problem and the channel-limitation task are investigated, where the proposed graph-learning-based parameter predictor enhances QAOA performance on both tasks, improving both the approximation ratio and computational efficiency. Furthermore, due to the complexity of the channel-limitation task and the scarcity of its pretraining data, a transfer learning strategy is employed to leverage knowledge from the original OPP task, where QAOA parameter datasets are more readily available to train the graph learning framework for the QAOA parameter predictor. The transfer learning approach also outperforms both random initialization and graph learning trained solely on the channel-limitation dataset in terms of the approximation ratio and the time efficiency. In general, this work is aimed at providing a new benchmark for solving complicated real-world power system optimization problems in the current NISQ era
Data supporting 'The effect of ambient and injection pressure on droplet size of ammonia sprays in a constant volume chamber'
No full textData supporting The Effect of Ambient and Injection Pressure on Droplet Size of Ammonia Sprays in a Constant Volume Chamber
Lipoprotein(a) is associated with coronary inflammation in people with HIV and undetectable HIV RNA
No full textAims: People with HIV (PWH) and undetectable virus experience elevated cardiovascular risk independent of traditional risk factors. Vascular inflammation may contribute to this residual risk. The perivascular fat attenuation index (FAI), derived from coronary computed tomography angiography (CCTA), is a biomarker of coronary inflammation. Lipoprotein(a) [Lp(a)] carries oxidized phospholipids that may promote inflammation. Statins have demonstrated cardiovascular benefit in PWH, including pleiotropic anti-inflammatory effects. This study assessed the associations of Lp(a) and of statin use with coronary inflammation (FAI) in men with HIV (MWH). Methods and results: We analysed FAI of the left anterior descending (LAD) and the right coronary arteries (RCA) in 583 men from the Multicenter AIDS Cohort Study, a prospective, multicentre cohort study, including 280 with undetectable HIV RNA, <50 copies/ml. Associations between log10[Lp(a)] and LAD and RCA FAI were assessed using linear regression, adjusting for demographic and cardiovascular risk factors. Log10[Lp(a)] was associated with LAD FAI in MWH with undetectable HIV in adjusted analysis [+1.99 HU (0.38, 3.59); P = 0.02] but not among men without HIV (MWoH) or MWH with detectable HIV. Associations with RCA FAI were only significant in the unadjusted analysis. Statin use was associated with lower FAI, less inflammation in the LAD in MWH with undetectable virus, but did not modify the association between Lp(a) and coronary inflammation. Conclusion: Lp(a) was associated with increased coronary inflammation, independent of traditional cardiovascular risk factors, in MWH with undetectable virus. Statin therapy did not modify the relationship between coronary inflammation and Lp(a)
Optimized bacterial expression of a synthetic BRIL antibody
No full textThe use of monoclonal fragments antigen binding (Fabs) is a prevalent methodology facilitating protein structure determination via both crystallography and cryo‐EM. The development of a synthetic Fab against the BRIL domain improved the accessibility of this approach, providing a general fiducial applicable to any protein of interest via the simple curation of a BRIL fusion protein. Here, we document the generation of a T7 Express ΔcybC strain allowing contaminant‐free bacterial expression of the synthetic anti‐BRIL Fab BAG2. We also report the crystal structure of BAG2 in complex with native cytochrome b562, a complex arising from expression in canonical Escherichia coli strains
Rendering White nationalism defensible in interaction
No full textPeople work to present themselves as moral, reasonable, or justified even when making racist or hateful comments. In this project, we identify interactional practices that accomplish support for White nationalism as part of a reasonable (or even positive) identity held by reasonable or positive actors. Using membership categorisation analysis and conversation analysis, we analysed a corpus of 24 publicly-available video recordings for explicit mentions of (or challenges to) White nationalism/supremacy. Looking for how people explain, justify, and rationalise White nationalism and, especially, White nationalist violence, we identified 16 cases of what we call White nationalist remediation. Our findings demonstrate how not only self-avowed White nationalists but also those who do not publicly identify as such can work to protect and potentially normalise White nationalist views and actions, including violence, using the same practices. They thus (1) signal their followers, (2) present their beliefs as reasonable and defensible, and (3) ultimately normalise White nationalism
Promoting women to managerial roles in the Bangladeshi garment sector
No full textWomen remain disadvantaged in promotion to managerial positions. We conduct a field experiment with 24 large garment factories in Bangladesh to test for inefficient representation of women among line supervisors. We identify the marginal female and male candidates for supervisory positions and randomly assign them to manage production lines. We document four findings: (1) In contrast to widespread negative beliefs about women’s ability as supervisors at baseline, female candidates selected by the factories had similar skills to males; (2) during the trial, females performed worse than males, which we show is related to negative bias against them; (3) after the trial, however, many female candidates were retained as supervisors and, conditional on that, performed similarly to males; and (4) after the end of our intervention, factories permanently increased the share of women among newly appointed supervisors. A conceptual framework of experimentation over discrimination rationalizes all these facts and cautions against the standard logic to test for discrimination: when there is uncertainty about the performance of the discriminated group, equal – or even worse – performance of the marginal candidates of that group is no longer sufficient to rule out inefficient discrimination
isMap – immunological synapse map analysis program
No full textT cell activation is initiated when T cells recognize their cognate antigen on the surface of antigen presenting cells (APCs). This triggers formation of a specialized membrane interface termed the immunological synapse (IS) which governs the spatial organization of the intercellular protein interactions ultimately determining the T cell response. While this is a fundamental process in adaptive immunity, tools for quantitative analysis and visualization of the IS molecular architecture are lacking. Here we present isMap, a computational framework for automated cell segmentation and quantification of various parameters related to T cell activation and IS formation on supported lipid bilayers (SLBs), including fluorescence intensity measurements, colocalization analysis and radial averaging. We validate isMap by confirming previous results showing that CD58 initially clusters with T cell receptor (TCR) before segregating into a distal ring during synapse maturation in activated CD8+ T cells. We also show that PD-L1 is initially distributed across the IS before ultimately accumulating with TCR in the center of the fully mature synapse. ICOSL, CD80 and CD86 cluster in the center of the contact area through all stages of IS maturation and colocalize with TCR in the order of ICOSL>CD86>CD80. These findings demonstrate isMap’s utility in dissecting the functional organization of the IS and highlight the dynamic redistribution of ligand-receptor pairs during T cell activation
Unlocking in-context learning for natural datasets beyond language modelling
No full textLarge Language Models (LLMs) exhibit In-Context Learning (ICL), which enables the model to perform new tasks conditioning only on the examples provided in the context without updating the model’s weights. While ICL offers fast adaptation across natural language tasks and domains, its emergence is less straightforward for modalities beyond text. In this work, we systematically uncover properties present in LLMs that support the emergence of ICL for autoregressive models and various modalities by promoting the learning of the needed mechanisms for ICL. We identify exact token repetitions in the training data sequences as an important factor for ICL. Such repetitions further improve stability and reduce transiency in ICL performance. Moreover, we emphasise the significance of training task difficulty for the emergence of ICL. Finally, by applying our novel insights on ICL emergence, we unlock ICL capabilities for various visual datasets and a more challenging EEG classification task. Code is available at https://github.com/jelenab98/unlocking_icl
