Among 143 cases of National Healthcare Safety Network (NHSN) catheter-associated urinary tract infections (CAUTI), 40% were considered catheter-associated asymptomatic bacteriuria (CA-ASB), and 18% clinical CAUTI. An alternative source of fever was present in 70% of CA-ASB. NHSN CAUTI may not be an effective metric for tracking hospital-level infection prevention efforts.

Text classification methods have been widely investigated as a way to detect content of low credibility: fake news, social media bots, propaganda, etc. Quite accurate models (likely based on deep neural networks) help in moderating public electronic platforms and often cause content creators to face rejection of their submissions or removal of already published texts. Having the incentive to evade further detection, content creators try to come up with a slightly modified version of the text (known as an attack with an adversarial example) that exploit the weaknesses of classifiers and result in a different output. Here we systematically test the robustness of common text classifiers against available attacking techniques and discover that, indeed, meaning-preserving changes in input text can mislead the models. The approaches we test focus on finding vulnerable spans in text and replacing individual characters or words, taking into account the similarity between the original and replacement content. We also introduce BODEGA: a benchmark for testing both victim models and attack methods on four misinformation detection tasks in an evaluation framework designed to simulate real use cases of content moderation. The attacked tasks include (1) fact checking and detection of (2) hyperpartisan news, (3) propaganda, and (4) rumours. Our experimental results show that modern large language models are often more vulnerable to attacks than previous, smaller solutions, e.g. attacks on GEMMA being up to 27% more successful than those on BERT. Finally, we manually analyse a subset adversarial examples and check what kinds of modifications are used in successful attacks.

Trends in elemental enrichment with stellar age can give us a powerful avenue to identify thus far unexplained origin sites of the elements. We investigate stellar abundance trends using the GALAH DR3 high-resolution spectroscopic dataset of 6 234 solar-type stars. Our study explores the elemental abundance [X/Fe] of sodium (Na) with stellar age. We find a pronounced enrichment in [Na/Fe] at super solar metallicity (i.e. [Fe/H] $ \gt \,0$) in the old sequence of Milky Way disc stars, a trend demanding a deeper understanding of the underlying source(s) responsible for the nucleosynthesis. This progressive [Na/Fe] enrichment at the young end of the old sequence has essential implications for Galactic archaeology. In this work, we propose a novel selection technique for separating the Milky Way’s thick and thin disc stellar populations (i.e. old and young sequences) based on the observed [Na/Fe] rise of $\sim$0.1 dex for stars around 5–8 Gyr old. We also compare our selection method to the conventional [Mg/Fe] vs. [Fe/H] selection approach, and we find that our new Na-based selection method better disentangles the overlap between young- and old-sequence disc stars at these intermediate ages. This is especially true at super solar [Fe/H], where the [Mg/Fe] vs. [Fe/H] or [$\alpha$/Fe] vs. [Fe/H] separation approaches exhibit significant overlap. This new selection method should help us better understand the history of the formation of the Milky Way disc.

Self-supervised representation learning (SSRL) in computer vision relies heavily on simple image transformations such as random rotation, crops, or illumination to learn meaningful and invariant features. Despite acknowledged importance, there is a lack of comprehensive exploration of the impact of transformation choice in the literature. Our study delves into this relationship, specifically focusing on microscopy imaging with subtle cell phenotype differences. We reveal that transformation design acts as a form of either unwanted or beneficial supervision, impacting feature clustering and representation relevance. Importantly, these effects vary based on class labels in a supervised dataset. In microscopy images, transformation design significantly influences the representation, introducing imperceptible yet strong biases. We demonstrate that strategic transformation selection, based on desired feature invariance, drastically improves classification performance and representation quality, even with limited training samples.

Fast and efficient identification is critical for reducing the likelihood of weed establishment and for appropriately managing established weeds. Traditional identification tools require either knowledge of technical morphological terminology or time-consuming image matching by the user. In recent years, deep learning computer vision models have become mature enough to enable automatic identification. The major remaining bottlenecks are the availability of a sufficient number of high-quality, reliably identified training images and the user-friendly, mobile operationalization of the technology. Here, we present the first weed identification and reporting app and website for all of Australia. It includes an image classification model covering more than 400 species of weeds and some Australian native relatives, with a focus on emerging biosecurity threats and spreading weeds that can still be eradicated or contained. It links the user to additional information provided by state and territory governments, flags species that are locally reportable or notifiable, and allows the creation of observation records in a central database. State and local weed officers can create notification profiles to be alerted of relevant weed observations in their area. We discuss the background of the WeedScan project, the approach taken in design and software development, the photo library used for training the WeedScan image classifier, the model itself and its accuracy, and technical challenges and how these were overcome.

As labor in the capitalist system practically tripled to some three billion workers, solidary organizations of labor simultaneously dwindled in relative size and power. This is true globally but also for the historical core countries. While this is a paradox, it is not a contradiction. Capital is a (spatialized) social relationship. The globalization of capital since the 1970s has shifted the power relations with localized labor fundamentally in favor of capital, as Charles Tilly noted in this journal almost thirty years ago. Over time, power balances within capitalist states, and between capitalist states and transnationalizing capital, have reflected that basic class-relational shift. This article explains why the globalizing cycle of weakened labor may now be reversing.

We report the non-breeding range of an adult Rapa Shearwater Puffinus myrtae, as estimated from data collected by one light logger deployed from 31 August 2019 to 22 July 2020. The Rapa Shearwater is classified as “Critically Endangered”, with a strong decline in breeding numbers reported recently. As the species is threatened by various introduced mammals on the breeding colonies, the main objective of this tracking essay was to identify the oceanic regions where the birds forage during the non-breeding season. The non-breeding range of this bird was located south-east of Rapa Island, where fishery activities are limited. The conservation efforts for the Critically Endangered Rapa Shearwater should first focus on securing mammal-free sites for breeding colonies.

This paper investigates the take-off performance of a single engine battery-electric aeroplane, using the example of the 300kg Sherwood eKub. It shows analysis of take-off performance of such an aeroplane must include as a minimum two new parameters not normally considered: time at full throttle and state of charge. It was shown in both ground and flight test that the state of available power reduces both as the throttle is fully open, and as battery charge is consumed, although recovers partially when power is reduced for a period. It is possible to schedule take-off performance as a function of the usual parameters plus state of charge. Because of the reducing climb performance with use of state of charge, and the requirement in airworthiness standards for minimum climb performance being available, it becomes necessary to introduce the concept of minimum-indicated state of charge for take-off, SoCiMTO; means to calculate that are shown for compliance with both microlight aeroplane standards and larger aeroplane standards, and the calculations are demonstrated for the eKub. Conclusions are also drawn about the use of commercial products SkyDemon and Google Earth for recording and analysing aeroplane performance data.

The One Health High-Level Expert Panel’s definition of One Health includes optimizing the health of people, animals (wild and domestic) and ecosystems. For many One Health practitioners, wildlife that can spread zoonoses are the focus, particularly if they can come in contact with people. However, ecosystem health is often best-indicated by less-encountered species, for instance, amphibians and reptiles. This review highlights how these taxa can benefit human health and well-being, including cultural significance, as well as their impact on plant, animal and environmental health. We highlight current challenges to the health of these species and the need to include them in the One Health Joint Action Plan. We conclude with a call to action for inclusion of amphibians and reptiles in a One Health approach.

Discriminatory morpho-metric features are obvious on legume seeds. This study utilized seven quantitative and 11 qualitative seed traits to characterize 139 African yam bean (AYB) breeding lines which were developed through single seed descent procedure. The seven quantitative data were subjected to analysis of variance, their means were combined with qualitative scores for genetic distance, principal component (PC) and clustering analyses. Significant (P ≤ 0.001) variation existed among the breeding lines for the seven traits. Mean ranges of seed length (SL), width (SW), thickness (ST) and a single seed weight (SSW) among the 139 breeding lines were respectively: 6.77–10.22 mm, 5.70–7.86 mm, 4.96–7.45 mm and 0.15–0.42 g. Positive and significant (P ≤ 0.05) genotypic correlation existed among SSW, SL, SW and ST. Seed colours, pattern, shapes, sizes, surface texture, brilliance varied among the breeding lines. Ranges of phenotypic and genotypic coefficient of variation and broadsense heritability were: 5.49–23.84%, 2.95–19.88% and 28.91–69.54% respectively. Fourteen (quantitative and qualitative) traits contributed higher (≥ 0.30) eigenvector loadings to the first three PC axes which explained 57.9% of the total variation among the breeding lines. Similarity among the lines was 0.75. Four clusters ensued in the dendrograph and each group had genetic similarities of: 0.85 (I), 0.82 (II), 0.78 (III) and 0.80 (IV). This research unveiled significant variation among AYB breeding lines with promising reliability for breeding opportunities of the qualitative and quantitative seed traits, which could contribute to higher grain yield and acceptability.

This paper analyses Italian party positions on the EU's response to the Russo-Ukrainian war, singling out the adoption of sanctions against Russia, the provision of military support to Kiev, enlargement to Ukraine and the welcoming of Ukrainian refugees into the Union's territory as the four main dimension of such a response. The paper draws on the literatures on cleavage politics, the inverted U curve and the differentiated forms of politicisation, thereby testing theory-driven research hypotheses through a qualitative content analysis of Italian parties' Facebook posts in the three months following the outbreak of the conflict, combining an inductive and a deductive approach. The findings show that party families are a good explanatory factor behind Italian party positions vis-à-vis the EU's response to the war outbreak as parties belonging to the same family shared a similar stance on the four dimensions of such a response. On the contrary, the Europeanism/Euroscepticism divide does not explain Italian party positions on the EU's reaction to the Ukrainian conflict as Europeanist parties split over the EU's provision of weapons to Ukraine about as much as Eurosceptic partis split over the adoption of sanctions against Moscow. Finally, the paper shows that policy issues in the EU's response to the war (such as sanctions and arms delivery) were much more salient for and contested by Italian political parties than constitutive issues (such as enlargement and asylum).