192815 research outputs found
Sort by
Effect of sacubitril/valsartan in heart failure with preserved ejection fraction across the age spectrum in PARAGON‐HF
Get PDFAims:
To evaluate clinical outcomes, echocardiographic features, and the efficacy and safety of sacubitril/valsartan compared to valsartan across age groups in the PARAGON-HF trial.
Methods and results:
A total of 4796 participants ≥50 years of age with chronic heart failure (HF) and left ventricular ejection fraction (LVEF) ≥45% were divided into three age groups: <65 years (n = 825), 65–74 years (n = 1772), and ≥75 years (n = 2199). Echocardiograms of 1097 patients were analysed in a standardized fashion at a core imaging laboratory. The primary composite outcome was total HF hospitalizations and cardiovascular (CV) death. Older patients were more likely to experience primary composite outcomes (compared to patients <65 years, adjusted rate ratio [aRR] for ≥75 years: 1.39, 95% confidence interval [CI] 1.21–1.61), total HF hospitalization (aRR 1.27, 95% CI 1.09–1.49), and CV death (adjusted hazard ratio [aHR] 2.04, 95% CI 1.44–2.87). Age did not modify the effect of sacubitril/valsartan compared to valsartan on primary composite endpoint (pinteraction = 0.79) in the overall population or in those with LVEF ≤57%. Older adults randomized to sacubitril/valsartan were more likely to develop hypotension compared to those receiving valsartan (pinteraction = 0.026). Older patients had smaller left ventricular chamber sizes, higher LVEF, and were more likely to have abnormal measures of diastolic function.
Conclusion:
Older patients with HF with preserved ejection fraction had higher event rates than younger patients, more adverse events overall, and more hypotension when treated with sacubitril/valsartan; however, the treatment benefits of sacubitril/valsartan were retained in older patients
Exploring the Effectiveness of Multi-stage Fine-tuning for Cross-encoder Re-rankers
No full textState-of-the-art cross-encoders can be fine-tuned to be highly effective in passage re-ranking. The typical fine-tuning process of cross-encoders as re-rankers requires large amounts of manually labelled data, a contrastive learning objective, and a set of heuristically sampled negatives. An alternative recent approach for fine-tuning instead involves teaching the model to mimic the rankings of a highly effective large language model using a distillation objective. These fine-tuning strategies can be applied either individually, or in sequence. In this work, we systematically investigate the effectiveness of point-wise cross-encoders when fine-tuned independently in a single stage, or sequentially in two stages. Our experiments show that the effectiveness of point-wise cross-encoders fine-tuned using contrastive learning is indeed on par with that of models fine-tuned with multi-stage approaches. Code is available for reproduction at https://github.com/fpezzuti/multistage-finetuning
Improving Low-Resource Retrieval Effectiveness Using Zero-Shot Linguistic Similarity Transfer
No full textGlobalisation and colonisation have led the vast majority of the world to use only a fraction of languages, such as English and French, to communicate, excluding many others. This has severely affected the survivability of many now-deemed vulnerable or endangered languages, such as Occitan and Sicilian. These languages often share some characteristics, such as elements of their grammar and lexicon, with other high-resource languages, e.g. French or Italian. They can be clustered into groups of language varieties with various degrees of mutual intelligibility. Current search systems are not usually trained on many of these low-resource varieties, leading search users to express their needs in a high-resource language instead. This problem is further complicated when most information content is expressed in a high-resource language, inhibiting even more retrieval in low-resource languages. We show that current search systems are not robust across language varieties, severely affecting retrieval effectiveness. Therefore, it would be desirable for these systems to leverage the capabilities of neural models to bridge the differences between these varieties. This can allow users to express their needs in their low-resource variety and retrieve the most relevant documents in a high-resource one. To address this, we propose fine-tuning neural rankers on pairs of language varieties, thereby exposing them to their linguistic similarities (https://github.com/andreaschari/linguistic-transfer). We find that this approach improves the performance of the varieties upon which the models were directly trained, thereby regularising these models to generalise and perform better even on unseen language variety pairs. We also explore whether this approach can transfer across language families and observe mixed results that open doors for future research
Generational gatekeepers: parental impediments to Global Englishes Language Teaching
No full textToday, the majority of students learn English to use as a global lingua franca. New needs require revised curricula. Global Englishes Language Teaching (GELT) responds to this need for innovation to reflect the current use of English as a Lingua Franca (ELF). However, innovation is a complex matter, particularly when it involves divergence from ‘standard’, ‘native’ English norms. This article explores the role of parents in the implementation of GELT, an under-researched group in the growing body of research. Interviews (n = 21) conducted before and after a GELT course on a postgraduate Teaching English to Speakers of Other Languages (TESOL) programme in the United Kingdom, plus six post-course focus groups (n = 27) explored pre- and in-service TESOL practitioners’ attitudes toward GELT. Our study explores this data to examine to what extent these teachers considered parental influence had on curriculum innovation. Findings reveal perceptions that parents still have strong attachments to ‘standard’ English and native-speakerism, hindering potential GELT implementation and particularly on altering teacher hiring practices. A growing number of studies focus on GELT awareness-raising amongst students and teachers. Our study highlights the need for more research on parental attitudes with regard to GELT, as well as how to raise their awareness of modern linguistic realities and how to ultimately involve them in the GELT innovation process
Moral enhancement and cheapened achievement: psychedelics, virtual reality and AI
Get PDFA prominent critique of cognitive or athletic enhancement claims that certain performance-improving drugs or technologies may ‘cheapen’ resulting achievements. Considerably less attention has been paid to the impact of enhancement on the value of moral achievements. Would the use of moral enhancement (bio)technologies, rather than (solely) ‘traditional’ means of moral development like schooling and socialization, cheapen the ‘achievement’ of morally improving oneself? We argue that, to the extent that the ‘cheapened achievement’ objection succeeds in the domains of cognitive or athletic enhancement, it could plausibly also succeed in the domain of moral enhancement—but only regarding certain forms. Specifically, although the value of moral self-improvement may be diminished by some of the more speculative and impractical forms of moral enhancement proposed in the literature, this worry has less force when applied to more plausibly viable forms of moral enhancement: forms in which drugs or technologies play an adjunctive or facilitative, rather than a determinative, role in moral improvement. We illustrate this idea with three examples from recent literature: the possible use of psychedelic drugs in certain moral-learning contexts, ‘Socratic Al’ (a proposed Al-driven moral enhancer) and empathy enhancement through virtual reality (VR). We argue that if one assumes that these technologies work roughly as advertised, the ‘cheapened achievement’ objection loses much of its bite. The takeaway lesson is that moral enhancement in its most promising and practical forms ultimately evades a leading critique of cognitive and athletic enhancement. We end by reflecting on the potential upshot of our analysis for enhancement debates more widely
Electrical phenomena in trees and wood: a review
Get PDFPurpose of review: This review covers electrical phenomena originating from the physical properties of wood, relevant to tree biology and timber industry applications. Membrane-associated cellular phenomena like action potentials are excluded. Trees exhibit diverse bioelectric processes of physical origin. The electrical properties of wood hold promise for advancing timber processing, and developing smart materials, while enhancing our understanding of tree-environment interactions.
Recent findings: Streaming and piezoelectric potentials have long histories but are now reinterpreted based on our better understanding of plants. Streaming potentials from sap flow, discounted in the 2000s, have been reinstated through recent publications addressing past inconsistencies with current data on xylem structure. Electro-osmotic flow is gaining new applications in timber drying. Wood, previously considered weakly piezoelectric, shows much stronger activity after fungal
degradation, spurring interest in practical applications and the underlying mechanism – now better understood through new findings on the structure, deposition and deformation of wood cellulose. Internal variation in the electric (conductive and dielectric) properties of green logs facilitates innovative timber quality mapping methods. Emerging research on perturbation of the atmospheric and soil electric fields by trees offers insights into inter-organism interactions.
Summary: This review encompasses electrical measurement methods; electrokinetic phenomena, including streaming potentials and electro-osmotic timber drying; electric heating; mapping and technologies based on dielectric properties; wood-based electronics; electromechanical phenomena, including the piezoelectric effect and triboelectrifcation; atmospheric electricity around trees; and electrotaxis. Future research should explore electro-osmosis in wood and its applications. Electric potentials in green wood and living trees, generated through ion-transport mechanisms, need further exploration to elucidate charge separation processes
The simpler the better? Threshold effects of energy labels on property prices and energy-efficiency investments
Get PDFUsing British data on over 7 million residential-property sales, we find that the provision of simplified energy-efficiency (EE) information leads to important market effects. UK legislation requires the use of an energy label that shows an EE rating score (1–100) mapped into arbitrary color-coded rating bands (A–G). Applying a regression-discontinuity design, we document significant price discontinuities at rating-band thresholds, providing evidence of heuristic thinking in the market. Importantly, these price discontinuities influence market behavior as sellers are more likely to make EE investments if their property is just below a threshold. We discuss policy considerations and avenues for future research
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇