University of St Andrews Research Portal
Not a member yet
81329 research outputs found
Sort by
Discriminating the Belarusian language in Belarus after 1995
No full textThe unconstitutional referendum of 1995 made Russian equal toBelarusian in its legal and administrative status in post-Soviet Belarus. As a result, the brief period of official monolingualism (also known as neo-Belarusianization) ended. The subsequent installation of dictatorship in Belarus took place in lockstep with the curbing of remaining public uses of Belarusian in favor of Russian. The final blow came in the wake of the 2020 mass protest and pro-democracy movement, sparked by the blatantly falsified presidential election this year. Brutal repressions followed, once again in synch with state-led actions aimed against organizations and publishers specializing in the support for and development of Belarusian language and culture. At present, the regime tends to openly see the Belarusian language as an indicator of ‘extremism,’meaning democratic and liberal values.Vujović, Novica, ed. 2025. Cetinjski filološki dani IV. Zbornik radova s međunarodnoga naučnog simpozijuma održanog na Cetinju 6–8. septembra 2023. Cetinje and Lawrence: Fakultet za crnogorski jezik i književnost and Department of Slavic, German & Eurasian Studies, University of Kansas.ISBN 978-9940-40-093-
Elizabeth L'Estrange, <i>Anne de Graville and Women's Literary Networks in Early Modern France</i> (Cambridge: D. S. Brewer, 2023)
No full textDevelopment and innovation in a new distributed medical programme:Scottish Graduate Entry Medicine (ScotGEM)
No full textIntroduction: Addressing the shortage of primary-care physicians, especially in remote and rural areas, is a crucial target in many countries. This article introduces the Scottish Graduate Entry Medicine (ScotGEM) programme: a compressed, tailor-made curriculum designed to equip and enthuse its graduates to practice generalist and rural medicine in Scotland, within the ethos of socially accountable medicine.Methods: This curriculum paper describes ScotGEM in sufficient detail for the reader to translate elements to their own context. It then collates findings from evaluations, research projects and many critical discussions about the programme. This work is used to describe and evaluate the curriculum design and delivery, with a focus on the distributed aspects.Results: Three key innovations of the curriculum are explored in detail: the Generalist Clinical Mentor (GCM) role; the year-long primary care Longitudinal Integrated Clerkship (LIC); and the Agents of Change curriculum. There are early signs that ScotGEM is encouraging generalist, rural careers within Scotland. There is also growing evidence of the benefits ScotGEM faculty and students bring to the clinical workforce in the distributed settings.Discussion: Distributed programmes require additional organization for students and faculty. Partnerships can be challenging but immensely rewarding. Healthcare partners in rural areas need to be involved early in planning and strong relationships fostered with local “champions.
Sample size considerations for species co-occurrence models
No full textMultispecies occupancy models are widely applied to infer interactions in the occurrence of different species, but convergence and estimation issues under realistic sample sizes are common. We conducted a simulation study to evaluate the ability of a recently developed model to recover co-occurrence estimates under varying sample size and interaction scenarios while increasing model complexity in two dimensions: the number of interacting species and the number of covariates. Using both standard and penalized likelihood, we demonstrate that the ability to quantify interactions in species occupancy using this model is highly sensitive to sample size, detection probability, and interaction strength. In the simplest scenario, there is high bias in the interaction parameter (used for co-occurrence inference) with less than 100 sites at high detection, and 400–1000 sites at low detection, depending on interaction strength. Strong co-occurrence is detected consistently above 200 sites with high detection probabilities, but weak co-occurrence is never consistently detected even with 2980 sites. We demonstrate that the mean predictive ability of the co-occurrence model is less affected by sample size, with low bias in derived probabilities at 50 sites. Our results highlight that while occupancy patterns are often robust to sample size limitations, reliable inference about co-occurrence demands substantially larger datasets than many studies currently achieve. We caution the interpretation of model output in small datasets or when co-occurrence is expected to be weak, but show methods are suitable to quantify strong co-occurrence in larger datasets and generate predictions of site occupancy states
Ted Hughes's <i>Tales from Ovid</i>:translation as a poetic dialogue with the Roman poet
No full textAspects of the commuting graph
No full textWe discuss the computational problem of deciding whether a given graphis the commuting graph of a finite group; we give a quasipolynomial algorithm,and a polynomial algorithm for the case when the group is an extraspecialp-group for p< an odd prime.We give new results on the question of whether the commuting graph ofa given group is a cograph or a chordal graph, two classes of graphs definedby forbidden subgraphs.The problems are not unrelated, since there are a number of cases where hardcomputational problems on graphs are easier when restricted to specialclasses of graphs; we conjecture that the recognition problem is polynomialfor cographs and chordal graphs
Do Large Language Models Have a Planning Theory of Mind?:Evidence from MindGames: a Multi-Step Persuasion Task
No full textRecent evidence suggests Large Language Models (LLMs) display Theory of Mind (ToM) abilities. Most ToM experiments place participants in a spectatorial role, wherein they predict and interpret other agents' behavior. However, human ToM also contributes to dynamically planning action and strategically intervening on others' mental states. We present MindGames: a novel `planning theory of mind' (PToM) task which requires agents to infer an interlocutor's beliefs and desires to persuade them to alter their behavior. Unlike previous evaluations, we explicitly evaluate use cases of ToM. We find that humans significantly outperform o1-preview (an LLM) at our PToM task (11% higher; ). We hypothesize this is because humans have an implicit causal model of other agents (e.g., they know, as our task requires, to ask about people's preferences). In contrast, o1-preview outperforms humans in a baseline condition which requires a similar amount of planning but minimal mental state inferences (e.g., o1-preview is better than humans at planning when already given someone's preferences). These results suggest a significant gap between human-like social reasoning and LLM abilities
Evaluating the galaxy formation histories predicted by a neural network in pure dark matter simulations
No full textWe investigate a series of galaxy properties computed using the merger trees and environmental histories from dark-matter–only cosmological simulations, using a semirecurrent neural network producing self-consistent predictions of galaxy evolution, and using stochastic improvements to this model based on similarly predicted Fourier transforms. We apply these methods to the dark-matter–only runs of the IllustrisTNG simulations to understand the effects of baryon removal, and to the gigaparsec-volume pure dark matter simulation Uchuu, to understand the effects of the lower resolution or alternative metrics for halo properties. We find that the machine learning model recovers accurate summary statistics derived from the predicted star formation and stellar metallicity histories, and correspondent spectroscopy and photometry. However, the inaccuracies of the model’s application to dark simulations are substantial for low-mass and slowly growing haloes. For these objects, the halo mass accretion rate is exaggerated due to the lack of stellar feedback, yet the formation of the halo can be severely limited by the absence of low-mass progenitors in a low-resolution simulation. Furthermore, differences in the structure and environment of higher mass haloes results in an overabundance of red, quenched galaxies. These results signify progress towards a machine learning model which builds high fidelity mocks based on a physical interpretation of the galaxy–halo connection, yet they illustrate the need to account for differences in halo properties and the resolution of the simulation
Determining the impact of histology on the incidence, pattern, and timing of recurrences in patients with renal cell carcinoma:a pooled analysis from the SORCE and ASSURE Trials
No full textBackground and objective: Outcomes after nephrectomy for intermediate- and high-risk renal cell carcinoma (RCC) according to histological subtype are poorly characterised. This study aims to determine the value of RCC histology in predicting survival and to inform on surveillance strategies in relation to patterns of first recurrence.0Methods: We pooled data from phase 3 trials: SORCE (n = 1689) and ASSURE (n = 1853). Of 3542 patients, 2881 had clear-cell RCC (ccRCC), 269 had papillary RCC (pRCC), 201 had chromophobe RCC (chRCC), and 191 had sarcomatoid RCC (sRCC). Relapse rates, median time to relapse (TTR), and first relapse sites were reported. Multivariable Cox regression models evaluated overall survival by histology, adjusting for initial relapse location and other important clinical factors. Key findings and limitations: Patients with pRCC and ccRCC had similar overall survival (log-rank p = 0.1). The median TTR for those with pRCC was 1.34 yr (interquartile range [IQR] 0.76, 2.59) compared with 1.78 yr (IQR 0.96, 3.38) for ccRCC patients (p = 0.012). Patients with chRCC had a median TTR of 2.72 yr (IQR 0.91, 4.11), and those with sRCC had a median TTR of 0.74 yr (IQR 0.50, 1.55). For sRCC patients, relapsing in the chest was associated with a lower risk of death than those relapsing in the abdomen (hazard ratio [HR] 0.5, confidence interval [CI]: 0.3, 0.88; p = 0.06). A similar trend was shown for pRCC (HR 0.5, CI: 0.2, 1.3; p = 0.1). Recurrence patterns for World Health Organization 2020 molecularly classified RCCs were not included. Despite pooling phase three datasets, small event numbers led to imprecise estimates, particularly for chRCC. Conclusions and clinical implications: Patients with intermediate and high-risk pRCC relapse earlier than those with ccRCC. Papillary RCC and sRCC first recurring in the abdomen exhibit poor prognosis, warranting consideration of additional abdominal imaging to enhance early relapse detection. ChRCC exhibits favourable prognosis and could avoid image-based surveillance until year 2. Patient summary This study evaluates pooled data from large phase 3 trials to precisely delineate relapse patterns for patients with intermediate- and high-risk cell renal cell carcinoma (RCC) according to their histology. The site and timing of first relapse provide useful information to support histology-specific RCC surveillance after nephrectomy. Development of genetic and molecular signatures corresponding to relapses at poor prognosis sites for each histology will individualise follow-up and is the next step
