Search

Cognitive scientists attribute representations to complex systems in order to explain their behavior. The shocking facility with which Large Language Models (LLMs) perform difficult linguistic and non-linguistic tasks has generated an increasing amount of speculation concerning what sorts of internal representations might underlie this behavior (whether personal, sub-personal, and of which kinds) and what properties such representations might have (for instance, whether they are grounded). This paper aims to elaborate and defend a conservative explanatory methodology, based on analyses of particular LLM behaviors, according to which attribution of sub-personal representations is key to explaining model performance which is robust, systematic, and flexible, especially in zero-shot settings, and that behavioral benchmarking alone is insufficient to resolve questions about representation due to the mutual underdetermination of performance and competence. The resulting view should help frame future explanations of LLM behavior, and provide an empirically grounded alternative to mere a priori speculation.

Objective. To compare the adequacy, agreement, and acceptability of Papanicolaou testing (cytology) for cervical cancer screening using self-collected samples compared to physician-collected samples in Grenada in the Caribbean. Furthermore, the study identifies the human papillomavirus (HPV) genotypes present among asymptomatic women testing positive for HPV, the etiologic cause of cervical cancer. Methods. Participants were divided into two groups and two cervical samples were collected from the women in each group: a self-collected sample and a physician-collected sample. Cervical specimens were tested for cytology and HPV. HPV genotyping was performed on positive specimens. Results. Self-collected samples were adequate and in agreement with physician-collected samples, showing no difference between the two sampling methods. Oncogenic high-risk HPV genotypes were identified in cervical samples which were positive for atypical squamous cells and low-grade squamous intraepithelial lesions. The high-risk HPV genotypes found, notably HPV 45 and 53, differed from those most commonly reported. Although the commonly reported high-risk genotypes HPV 16 and 18 were found, so were 31, 33, 35, 52, 66, 68, and 82. Conclusions. Using self-collection facilitated the discovery of unexpected HPV genotypes among asymptomatic women in Grenada. These findings add new information to the literature regarding cervical cancer and neoplasia screening and HPV genotypes in the Caribbean. This genotype information may impact surveillance of women with low-grade lesions, HPV vaccine selection, and possibly further vaccine research. Research regarding HPV in Caribbean pathology samples of cervical neoplasia and cancer is needed.

Background Mitral annular disjunction (MAD), posterior displacement of the mitral valve leaflet hinge point, predisposes to arrhythmias or sudden cardiac death. We evaluated the burden of MAD, mitral valve prolapse (MVP), and mitral regurgitation (MR) by heritable thoracic aortic disease gene in a cross‐sectional analysis of 2014–2023 data in the Montalcino Aortic Consortium registry. Methods and Results MAD was determined by direct measurement of echocardiographic images. MR and MVP were defined according to current clinical guidelines. Associations were evaluated using χ2 or Fisher exact tests. MR and MVP were enriched in Montalcino Aortic Consortium participants (672) with pathogenic variants (PV) in transforming growth factor‐β pathway genes. The combination of MR and MVP was associated with mitral surgery and arrhythmias. In the subgroup with available images, MAD was enriched in SMAD3 PV compared with other transforming growth factor‐β PV (prevalence ratio 1.8 [1.1–2.8], P 10 mm) was only observed in the transforming growth factor‐β subgroup and was further enriched in participants with SMAD3 PV (prevalence ratio 3.1 [1.1–8.6]). MVP (prevalence ratio 5.2 [3.0–9.0]) and MR (PR 2.7 [1.8–3.9]) were increased in participants with MAD, but MAD was not independently associated with adverse cardiac or aortic events. Conclusions Pathological mitral valve phenotypes are more prevalent in individuals with PV in transforming growth factor‐β pathway genes, particularly SMAD3. MR and MVP but not MAD are associated with adverse aortic and cardiac events. Because congenital mitral disease may be the primary presenting feature of SMAD3 PV, genetic testing for heritable thoracic aortic disease should be considered for such individuals, especially if they also have a family history of heritable thoracic aortic disease.

Even the most conscientious nonprofit organizations can fall victim to fraud. We examine how a nonprofit organization’s Form 990 disclosures and media coverage about an asset diversion influence subsequent donor support. Consistent with a loss in trust, we observe a decrease in donations following a diversion. This decrease is amplified when the diversion is reported in the news. We also find that donations decline more when organizations do not provide transparent disclosures and when losses are higher but only if the diversion receives media coverage. Finally, our results indicate that donors punish organizations less when they report higher recoveries and governance improvements. This study describes mechanisms though which news coverage enhances donor oversight. The media can: (a) directly inform some donors, (b) prompt some donors to obtain further information, and (c) motivate organizations to provide higher-quality disclosure.

Bicuspid aortic valve (BAV) is the most common congenital heart malformation in adults, but it can also cause childhood-onset complications. The presentation and clinical course of young adults who present due to BAV complications are relatively uncharacterized. In a multicenter study, we found that young people who experience significant complications related to BAV disease before age 30 are distinguished from the majority of BAV cases that manifest after age 50 by a relatively severe clinical course, with higher rates of surgical interventions, more frequent second interventions, and a greater burden of congenital heart malformations. These observations highlight the need for prompt recognition, regular lifelong surveillance, and targeted interventions to address the significant health burdens of patients with early-onset BAV complications.

BACKGROUND: The impact of adjunctive anti-inflammatory treatment on outcomes for patients with Kawasaki disease (KD) and coronary artery aneurysms (CAAs) is unknown. METHODS: Using data from the International KD Registry in patients with ≥ medium CAA we evaluate associations of treatment with outcomes and major adverse cardiac events (MACE). RESULTS: Medium or large CAA was present in 527 (32%) patients. All were treated with intravenous immunoglobulin (IVIG), 70% were male, and the median age was 1.3 years (interquartile range: 0.4-4.0 years). The most common acute therapies included single IVIG alone in 243 (46%), multiple IVIG in 100 (19%), multiple IVIG + corticosteroids in 75 (14%), and multiple IVIG + infliximab + corticosteroids in 44 (8%) patients. Patients who received therapy beyond single IVIG had a larger CA z-score at baseline (P < 0.001) and a higher rate of bilateral CAA (P < 0.001). Compared with IVIG alone, early adjunctive treatments (within 3 days of initial IVIG) were not associated with time to CAA regression or MACE, whereas later adjunctive therapy was associated with MACE and longer time to CAA regression. Patients receiving IVIG plus steroids vs IVIG alone had a trend towards shorter time to CAA regression and lower risk of MACE (P = 0.07). A larger CAA z-score at baseline was the strongest predictor of an increase in the CAA z-score over follow-up, lower likelihood of CAA regression, and higher risk of MACE. CONCLUSIONS: Persistence of CAA and MACE are more strongly associated with baseline severity CAA than with acute adjuvant anti-inflammatory therapy. Patients who received late adjunctive therapy are at higher risk for worse outcomes.

Citizen science programs offer opportunities for K-12 students to engage in authentic science inquiry. However, these programs often fall short of including learners as agents in the entire process, and thus contrast with the growing open science movement within scientific communities. Notably, study ideation and peer review, which are central to the making of science, are typically reserved for professional scientists. This study describes the implementation of an open science curriculum that engages high school students in a full cycle of scientific inquiry. We explored the focus and quality of students' study designs and peer reviews, and their perceptions of open science based on their participation in the program. Specifically, we implemented a human brain and behavior citizen science unit in 6 classrooms across 3 high schools. After learning about open science and citizen science, students (N = 104) participated in scientist-initiated research studies, and then collaboratively proposed their own studies to investigate personally interesting questions about human behavior and the brain. Students then peer reviewed proposals of students from other schools. Based on a qualitative and quantitative analysis of students' artifacts created in-unit and on a pre and posttest, we describe their interests, abilities, and self-reported experiences with study design and peer review. Our findings suggest that participation in open science in a human brain and behavior research context can engage students with critical aspects of experiment design, as well as with issues that are unique to human subjects research, such as research ethics. Meanwhile, the quality of students' study designs and reviews changed in notable, but mixed, ways: While students improved in justifying the importance of research studies, they did not improve in their abilities to align methods to their research questions. In terms of peer review, students generally reported that their peers' feedback was helpful, but our analysis showed that student reviewers struggled to articulate concrete recommendations for improvement. In light of these findings, we discuss the need for curricula that support the development of research and review abilities by building on students' interests, while also guiding students in transferring these abilities across a range of research foci.

IntroductionPhysical exercise and cognitive training have the potential to enhance cognitive function and mobility in older adults at risk of Alzheimer's disease and related dementia (ADRD), but little is known about the feasibility of delivering multidomain interventions in home settings of older adults at risk of ADRD. This study aims to assess the feasibility of home-based delivery of exercise and cognitive interventions, and to evaluate the relationship between participants' intervention preferences and their subsequent adherence. Secondary objectives include the effect of the interventions on ADRD risk factors, including frailty, mobility, sleep, diet and psychological health.Methods and analysisThe SYNchronising Exercises, Remedies in GaIt and Cognition at Home (SYNERGIC@Home) feasibility trial is a randomised control trial that follows a 2×2 factorial design, with a 16-week home-based intervention programme (3 sessions per week) of physical exercises and cognitive training. Participants will be randomised in blocks of four to one of the following four arms: (1) combined exercise (aerobic and resistance)+cognitive training (NEUROPEAK); (2) combined exercise+control cognitive training (web searching); (3) control exercise (balance and toning)+cognitive training; and (4) control exercise+control cognitive training. SYNERGIC@Home will be implemented through video conferencing. Baseline and post-intervention assessments at 4-month and 10-month follow-up will include measures of cognition, frailty, mobility, sleep, diet and psychological health. Primary feasibility outcome is adherence to the interventions. Primary analytic outcome is the relationship between pre-allocation preference for a given intervention and subsequent adherence to the allocated intervention. A series of secondary analytic outcomes examining the potential effect of the individual and combined interventions on cognitive, mobility and general well-being will be measured at baseline and follow-up.Ethics and disseminationEthics approval was granted by the relevant research ethics boards. Findings of the study will be presented to stakeholders and published in peer-reviewed journals and at provincial, national and international conferences.Trial registration numberNCT04997681, Pre-results.