150259 research outputs found
Sort by
Redirection for Erasing Memory (REM): Towards a universal unlearning method for corrupted data
No full textMachine unlearning is studied for a multitude of tasks, but specialization of unlearning methods to particular tasks has made their systematic comparison challenging. To address this issue, we propose a conceptual space to characterize diverse corrupted data unlearning tasks in vision classifiers. This space is described by two dimensions, the discovery rate (the fraction of the corrupted data that are known at unlearning time) and the statistical regularity of the corrupted data (from random exemplars to shared concepts). Methods proposed previously have been targeted at portions of this space and, as we show, fail predictably outside these regions. We propose Redirection for Erasing Memory (REM), whose key feature is that corrupted data are redirected to dedicated neurons introduced at unlearning time and then discarded or deactivated to suppress the influence of corrupted data. REM performs strongly across the space of tasks, in contrast to prior SOTA methods that fail outside the regions for which they were designed
Less is More: Pre-Training Cross-Lingual Small-Scale Language Models with Cognitively-Plausible Curriculum Learning Strategies
No full textCurriculum Learning has been a popular strategy to improve the cognitive plausibility of Small-Scale Language Models (SSLMs) in the BabyLM Challenge. However, it has not led to considerable improvements over non-curriculum models. We assess whether theoretical linguistic acquisition theories can be used to specify more fine-grained curriculum learning strategies, creating age-ordered corpora of Child-Directed Speech for four typologically distant language families to implement SSLMs and acquisition-inspired curricula cross-lingually. Comparing the success of three objective curricula (GROWING, INWARDS and MMM) that precisely replicate the predictions of acquisition theories on a standard SSLM architecture, we find fine-grained acquisition-inspired curricula can outperform non-curriculum baselines and performance benefits of curricula strategies in SSLMs can be derived by specifying fine-grained language-specific curricula that precisely replicate language acquisition theories
Complex de novo structural variants are an underestimated cause of rare disorders
No full textComplex de novo structural variants (dnSVs) are crucial genetic factors in rare disorders, yet their prevalence and characteristics in rare disorders remain poorly understood. Here, we conduct a comprehensive analysis of whole-genome sequencing data of 12,568 families, including 13,698 offspring with rare diseases, obtained as part of the UK 100,000 Genomes Project. We identify 1,870 dnSVs, constituting the largest dnSV dataset reported to date. Complex dnSVs (n = 158; 8.4%) emerge as the third most common type of SV, following simple deletions and duplications. We classify 65% of these complex dnSVs into 11 subtypes. Among probands with dnSVs (n = 1,696), 9% exhibit exon-disrupting pathogenic dnSVs associated with the probands’ phenotype. Notably, 12% of exon-disrupting pathogenic dnSVs and 22% of de novo deletions or duplications previously identified by array-based or whole-exome sequencing methods are found to be complex dnSVs. We also find distinct genomic properties of de novo deletions depending on the parent of origin. This study highlights the importance of complex dnSVs in the cause of rare disorders and demonstrates the necessity of specific genomic analysis to avoid overlooking these variants
Antisense reduction in NADP-ME in the C4 species Flaveria bidentis alters stomatal sensitivity to intercellular [CO2].
No full textStomata have the crucial role of balancing the uptake of CO2 for photosynthesis and water loss via transpiration by adjusting their aperture. In C3 plants, coordination between photosynthesis and stomatal conductance can be disrupted by mutations in Calvin-Benson-Bassham (CBB) cycle expression, but it is not clear whether the disruption of the C4 carbon concentrating mechanism would have similar effects. In this study, it was hypothesized that perturbing the C4 cycle, resulting in reduced carbon flux into the bundle sheath cells and overall net CO2 assimilation rates, may alter stomatal regulation. To test this hypothesis, we performed gas exchange measurements on transgenic Flaveria bidentis plants carrying an antisense construct leading to reduced nicotinamide adenine dinucleotide phosphate (NADP)-malic enzyme (NADP-ME) activity. Stomatal conductance was unaltered in the antisense plants at ambient CO2 levels. However, Ci was significantly increased. When measurements were instead performed at common Ci, stomatal conductance was significantly higher in the antisense lines. These results demonstrate that disruption of the C4 cycle via reduced NADP-ME activity in F. bidentis leads to altered stomatal sensitivity to Ci
The association between early diagnosis of gestational diabetes and maternal-neonatal outcomes: a secondary analysis of the digest trial
Get PDFPurpose: To determine whether early gestational diabetes (GDM) differs from later GDM in maternal characteristics and perinatal outcomes. Methods: This is a secondary analysis of an energy-restricted dietary intervention in GDM (DiGest) randomized controlled trial. We compared maternal weight, glycemia, and pregnancy/neonatal outcomes between the early GDM (< 20 weeks, n = 118) and standard GDM diagnosis (21–28 weeks, n = 299) groups. Results: Early GDM was associated with higher antenatal (40 vs. 38 mmol/mol; p = 0.017) and postnatal HbA1c (38 vs. 36 mmol/mol; p = 0.002) and higher risk of diabetes/prediabetes postnatally (19 vs. 6%; p = 0.003). Despite higher medication requirements, perinatal outcomes did not differ. Lower gestational weight gain (2.5 vs. 5.3 kg, p = 0.003) and comparable glycemia at 36 weeks were found. Timing of diagnosis did not impact the effect of the DiGest intervention. Conclusion: Early GDM reflects more severe underlying hyperglycemia, but timely treatment and reduced gestational weight gain can offset adverse perinatal risks
Advancing Cancer Prevention Through Precision Prediction
No full textSummary:
Emerging knowledge, integrative frameworks, and multidisciplinary partnerships together create an unprecedented opportunity to advance cancer prevention through precision. By combining multimodal risk assessment with causal biological insight and dynamic models of early carcinogenesis, precision prevention can enable actionable, equitable, and acceptable preventive and interception strategies
Gendered risks to children and adolescents assessed by Child & Adolescent Mental Health Services (CAMHS): Perspectives from network analysis
Get PDFBACKGROUND
Early exposure to risk and adversity is a potent predictor of mental health difficulties. Though risks vary by gender, little attention was paid towards the associations both within risks and of risks across genders.
OBJECTIVE
We sought to identify networks of a wider range of risks (experiences and behaviors that might threaten the person's wellbeing and safety before the age of 18 years). And we aimed to have a better understanding of the specific risk configurations across genders and to develop potential clinical interventions.
Participants and setting
This study explores network structures of early risks among 45,210 children and adolescents (aged 5 to 18) from longitudinal data in the UK.
METHODS
Network analysis was applied to investigate the associations among risks and to identify the central risks across genders.
RESULTS
Stable connections across genders in different assessments of risks (e.g., risks of self-harm and suicide). Risks related to violence could be core risks in all networks. Some gender differences in the context of early risks are also identified. For example, substance misuse and exhibiting violent or offending behavior are more closely associated among the male children that took the Brief Risk Assessment.
CONCLUSIONS
Gendered associations between risks could be of value for both intervention and prevention. More attention should be paid to risks related to violence in clinical practice and policy making. Future study could record risks more precisely, utilize data from multiple time points and take more social-demographic factors into consideration to obtain integrated and comprehensive results
Preoperative Imaging for Cochlear Implantation: A Global Consensus.
Get PDFOBJECTIVE: Preoperative imaging is vital for cochlear implant surgeries, aiding diagnosis, and surgical planning. This study evaluated global practices and the value of preoperative imaging through an international survey. METHODS: A cross-sectional survey was conducted among international cochlear implantation experts using a 112-item questionnaire. The study explored imaging modalities, anatomical targets, evaluated parameters, and different imaging approaches' perceived value and risks. Participants were recruited from a global consortium of otolaryngology, otology, neurotology, and cochlear implant surgery professionals. RESULTS: Thirty-nine practitioners from 36 centers in 16 countries completed the survey (95.1% response rate). All used computed tomography (CT) and magnetic resonance imaging (MRI) for preoperative evaluation; MRI was deemed more valuable for diagnosis and candidacy assessment, while CT was preferred for surgical planning. Nearly half utilized additional imaging modalities, with functional MRI being the most common (20.5%). Additionally, 79.5% of respondents reported using image-based surgical planning software. DISCUSSION: Results show a universal reliance on CT and MRI for cochlear implant evaluations, with MRI aiding diagnosis and CT focusing on surgical planning. Advanced imaging techniques may emerge in specific clinical cases. IMPLICATIONS FOR PRACTICE: Modern imaging practices and their potential changes can enhance protocol development and improve preoperative evaluations, ultimately boosting patient safety and outcomes in cochlear implantation
JWST/MIRI coronagraphic search for planets in systems with gapped exoKuiper belts and proper-motion anomalies
Get PDFABSTRACT Over the past decade, ALMA has uncovered a range of substructures within exoKuiper belts, pointing to a population of undetected planets. With James Webb Space Telescope (JWST)’s sensitivity, we now have the opportunity to identify these planets thought to be responsible for the observed substructures in debris discs. We present Cycle 1 JWST/MIRI m coronagraphic observations of three exoKuiper belts that exhibit gaps in their radial structures: HD 92945, HD 107146, and HD 206893, to determine whether planets are responsible for carving these structures, as seen in our Solar system with the gas giants. We reduce the JWST/MIRI data using spaceklip, and introduce new routines to mitigate the Brighter-Fatter effect and persistence. We do not detect any planet candidates, and all detected objects in the field of view are consistent with background stars or galaxies. However, by combining JWST mass limits, archival observational constraints, and astrometric accelerations, we rule out a significant portion of planet parameter space, placing tight constraints on the planets possibly responsible for these gaps. To interpret these results, we explore multiple gap-carving scenarios in discs, either massless or with non-zero mass, including clearing by in situ planet(s), as well as shaping by inner planets through mean-motion or secular apsidal resonances. Finally, we conclude that the planets causing the proper-motion anomaly in these systems must reside within the inner 20 au
Immune-mediated protection and enhancement of dengue drives patterns of infant cases in Brazil.
No full textInfants represent a vulnerable population for severe outcomes from infectious diseases. For dengue virus, maternal antibodies can cause severe disease in infants aged 5-12 months through antibody-dependent enhancement, but the role of maternal immunity in neonates aged 0-30 days has not been quantified. Our study leverages the re-emerging setting of Brazil, where the proportion of mothers with anti-dengue antibodies varies across space and time. Infant dengue burden has risen 11-fold from 2000-2024, in line with expansion in the general population. We fit mechanistic models to infant dengue and severe dengue cases reported through surveillance to disentangle the competing roles of maternal immunity, transmission, and age on the risk of dengue and severe dengue in infants across the first year of life. We observed peaks in severe dengue risk in neonates and at 7-8 months of age. These two peaks were explained by competing mechanisms. Infants born to seropositive mothers had lower overall risk of dengue, but increased risk of severe disease from 5-12 months, than those born to seronegative mothers. As dengue incidence increases, and vaccination becomes available, maternal antibodies may decrease infant burden but increase the risk of severe disease