Vibrant Data (United States)

Automated evaluation of generative text-to-image models remains a challenging problem. Recent works have proposed using multimodal LLMs to judge the quality of images, but these works offer little insight into how multimodal LLMs make use of concepts relevant to humans, such as image style or composition, to generate their overall assessment. In this work, we study what attributes of an image--specifically aesthetics, lack of artifacts, anatomical accuracy, compositional correctness, object adherence, and style--are important for both LLMs and humans to make judgments on image quality. We first curate a dataset of human preferences using synthetically generated image pairs. We use inter-task correlation between each pair of image quality attributes to understand which attributes are related in making human judgments. Repeating the same analysis with LLMs, we find that the relationships between image quality attributes are much weaker. Finally, we study individual image quality attributes by generating synthetic datasets with a high degree of control for each axis. Humans are able to easily judge the quality of an image with respect to all of the specific image quality attributes (e.g. high vs. low aesthetic image), however we find that some attributes, such as anatomical accuracy, are much more difficult for multimodal LLMs to learn to judge. Taken together, these findings reveal interesting differences between how humans and multimodal LLMs perceive images.

Over the past decades, our understanding of the vital role microbes play in ecosystem processes has greatly expanded. However, we still have limited knowledge about how microbial communities interact with larger organisms. Many existing representations of microbial interactions are based on co‐occurrence patterns, which do not provide clear insights into trophic or non‐trophic relationships. In this study, we untangled trophic and non‐trophic interactions between macroscopic and microscopic organisms on a marine rocky shore. Five abundant mollusk grazers were selected, and their consumptive (grazing) and nonconsumptive (grazer pedal mucus) interactions with bacteria in biofilms were measured using 16S rRNA‐gene amplicon sequencing. While no significant effects on a commonly used measure of biofilm grazing (chlorophyll a concentration) were observed, detailed image analysis revealed that all grazers had a detrimental impact on biofilm cover. Moreover, different grazers exhibited distinct effects on various bacterial groups. Members of the Alteromonadaceae, Burkholderiaceae, Flavobacteriaceae, Halieaceae, Phycisphaeraceae, Rhodobacteraceae, Rickettsiaceae, Saprospiraceae, and Vibrionaceae families experienced positive trophic effects from specific grazers. In contrast, members of the Flavobacteriaceae, Pirellulaceae, Rhodobacteraceae, Rubritaleaceae, and Saprospiraceae families were negatively affected by trophic interactions with other grazers. Some members of the Gammaproteobacteria, Flavobacteriaceae, Ilumatobacteraceae, Pirellulaceae, Rickettsiales, Rhodobacteraceae, and Rubritaleaceae families exhibited non‐trophic positive interactions with specific grazers. Meanwhile, members of the Family DEV007 (Verrucomicrobiales), Flavobacteriaceae, Ilumatobacteraceae, Legionellaceae, Rickettsiales, Rhodobacteraceae, Saprospiraceae, and Xanthobacteraceae families exhibited non‐trophic negative interactions with particular grazers. Both trophic and non‐trophic interactions shift the microbial community toward enhanced recycling, energy efficiency, and stress resilience. Grazer activity, through biomass removal and exudates like pedal mucus, reduces photosynthetic groups like diatoms, halting dimethylsulfoniopropionate (DMSP) production and negatively impacting sulfur‐cycling bacteria and associated parasites. This research complements the ecological network of the intertidal rocky shore in central Chile and represents the first attempt to construct an interaction network between macroorganisms and bacteria. It reveals that the strength of trophic and non‐trophic interactions varies depending on the grazer and bacterial group involved. While some bacterial groups responded broadly, others showed specialized responses to specific macroorganisms. Overall, this study highlights the potential for integrating microbes into ecological networks, offering valuable insights methodologies for quantifying interactions across domains.

This paper systematically studies the cost impact of distributed in-memory caches on datacenter services. While memory used for these caches is often perceived to be expensive, we find that the resulting CPU savings from these in-memory caches far outweigh the cost of added memory. In fact, across a variety of both synthetic and production workloads, we find that adding distributed in-memory caches can lower total operating costs by 3 – 4×, even without considering their latency benefits. These cost savings can vary significantly across various architectures, such as storage layer caches, remote lookaside caches, and in-memory linked caches. We additionally evaluate cost for two emerging scenarios: caching rich application objects and strongly consistent cache. For the former, we find that caching application objects provides outsized benefits compared to their denormalized, key-value-style variants, up to 8× compared to reading from storage. For the latter, we observe that even a minimal version check for consistency can eliminate most of the cost benefits, calling for new designs for cost-effective consistent cache.

Landscape connectivity is often negatively impacted by road networks that fragment habitat and result in genetic and demographic consequences for wildlife. Existing roadway structures like bridges, culverts, and underpasses can facilitate connectivity and reduce the barrier effect of roads by providing less risky areas for animals to cross. Estimating areas of high wildlife movement near roads is beneficial for prioritizing transportation investments for wildlife. We used an omnidirectional circuit theory approach to model the movements of eight terrestrial mammal species across the state of Vermont, a forested region central to the globally important Northern Appalachian ecoregion. We combined expert-derived landscape resistance surfaces with wildlife occurrence data to develop species-specific connectivity models at statewide (23,873 km2, 30 m resolution) and roadway structure (100 m radius around 5,912 structures, 0.5 m resolution) scales. The flow of animal movement across the landscape, depicted as electrical current density, was highest for forest-obligate species along the forested, mid-elevation foothills of the Green Mountains in central Vermont and lowest in the agricultural Champlain Valley; however, for more urban- and agriculture-adapted species, flow was highest in developed areas and lower elevation valleys. Average current density was highest for black bear (Ursus americanus), and lowest for striped skunk (Mephitis mephitis) at the statewide scale and highest for raccoon (Procyon lotor) and lowest for moose (Alces alces) at the finer structure scale. Results at both scales revealed different patterns of expected animal movement that reflect the relative extent of connectivity. We then scored connectivity for each structure across all species by combining both scales using four different methods to capture a range of management interests. Rankings varied greatly depending on the method used, highlighting the need to clearly articulate objectives when scoring structures or other features in a landscape. Resistance, occupancy, and current maps also indicated the broad importance of intact forest for connectivity and may be particularly important for identifying priority regions for protection under Vermont’s Community Resilience and Biodiversity Protection Act that mandates protecting 50% of the state by 2050.

Objectives:To test the psychometric properties of several dementia care-related scales among Latinos in the US.Design:We leveraged secondary baseline data from a one-arm mHealth trial on dementia caregiver support. We included 100 responses for caregiver-focused scales and 88 responses for care recipient-focused scales. Scales included the Neuropsychiatric Inventory Questionnaire Severity and Distress scales, six-item Zarit Burden Inventory, Ten-item Center for Epidemiologic Studies Depression Scale, Geriatric Depression Inventory, Quality of Life in Alzheimer’s Disease, and Single-item Satisfaction With Life Scale. We calculated concurrent validity using Pearson and Spearman correlations and expected correlations amongst all variables in line with the Stress Process Framework. We calculated internal consistency reliability using Cronbach’s alpha.Results:All concurrent validity correlations followed the expected directionality, with 19/21 inter-scale correlations in the total sample reaching statistical significance (p<0.05), and 17/21 reaching at least a low correlation (0.3). Cronbach’s alpha ranged from 0.832 to 0.879 in all scales in the total sample.Conclusion:The English and Spanish caregiver-administered scales tested in this manuscript have good psychometric properties.Clinical Implications:The dementia care-related scales are now appropriately available for use among US Latinos in research and clinical contexts.

Nos centraremos en las realizaciones de la vibrante en posición posnuclear de hablantes de nivel de instrucción bajo, analizando asimismo la

Oxidative stress refers to the imbalance between the production of oxidant species and the body’s ability to quench them using antioxidants, favoring the rise in oxidant levels. This leads to the damage of cellular macromolecules such as lipids, DNA, RNA, and proteins. The body’s ability to manage oxidative stress and maintain it at an optimum level is crucial for overall health. Oxidative damage, if left unmitigated, contributes to the aging process characterized by the progressive deterioration of physiological functions and cellular structures. Understanding the mechanisms of oxidative stress along with its reliable quantification can enable consistency and comparability in clinical practice across diseases. While direct quantification of oxidant species in the body would be ideal for assessing oxidative stress, it is not feasible owing to their high reactivity, short half-life, and quantification challenges using conventional techniques. Quantifying oxidative damage products and antioxidants pose as appropriate markers, indicating the degree of oxidative stress in the body. This review comprehensively discusses the mechanism of generation of key oxidant species, their sources, the beneficial roles played by them at low levels and the detrimental effects exerted by their elevated levels. The review also provides insights into the effective quantification techniques for damage products of lipids nucleic acids, and proteins along with the endogenous and exogenous antioxidant markers. Effective quantification of oxidative stress may improve our understanding on the phenomenon which may aid in maintaining cellular integrity, preventing age-associated diseases, and thereby promoting optimal well-being and longevity.

This study aimed to examine how occupational therapists and students perceive systemic racism in occupational therapy practice and the effectiveness and impact of the delivery of a keynote address on confronting systemic racism in practice delivered to occupational therapists and students. The study aimed to help inform future efforts in creating a more diverse and inclusive profession at the practice, policy, and education level by providing input into the challenges and opportunities that occupational therapists and students face when reckoning with issues of racism in practice. A mixed methodology research design was used with Likert-style scale and open-ended questions delivered via online survey. Eighty-nine occupational therapists and students completed the online survey. Most of the respondents reported responding favorably to the keynote address and shared sentiment that it could be a difficult and uncomfortable topic, yet it was important to discuss because of its impact on the profession and the clients occupational therapists serve.

The densest subgraph problem has received significant attention, both in theory and in practice, due to its applications in problems such as community detection, social network analysis, and spam detection. Due to the high cost of obtaining exact solutions, much attention has focused on designing approximate densest subgraph algorithms. However, existing approaches are not able to scale to massive graphs with billions of edges. In this paper, we introduce a new framework that combines approximate densest subgraph algorithms with a pruning optimization. We design new parallel variants of the state-of-the-art sequential Greedy++ algorithm, and plug it into our framework in conjunction with a parallel pruning technique based on k-core decomposition to obtain parallel (1+ε)-approximate densest subgraph algorithms. On a single thread, our algorithms achieve 2.6-34× speedup over Greedy++, and obtain up to 22.37× self-relative parallel speedup on a 30core machine with two-way hyper-threading. Compared with the state-of-the-art parallel algorithm by Harb et al. [NeurIPS'22] , we achieve up to a 114× speedup on the same machine. Finally, against the recent sequential algorithm of Xu et al. [PACMMOD'23] , we achieve up to a 25.9× speedup. The scalability of our algorithms enables us to obtain near-optimal density statistics on the hyperlink2012 (with roughly 113 billion edges) and clueweb (with roughly 37 billion edges) graphs for the first time in the literature.

In data platforms with thousands of data tables available for exploration, users often need to retrieve some data based on limited knowledge of the data sources and schema. The task that automates retrieving attributes from an online data lake given a set of entities from users is called "entity augmentation". The key for successful entity augmentation is an accurate construction of semantic relationships between data tables. Current techniques either focus on retrieving categorical values or numerical values with pre-defined rules. Further, they assume there are meta-data available for each table, such as texts and tags. In this paper, we introduce a semantic graph for numerical data augmentation that (i) matches columns with similar semantic relationships without any meta-data from the tables; (ii) infer the conversion rules among different numerical columns based on the values. The approach is designed to be highly scalable and parallel for large-scale data lakes with millions of large datasets. We also propose efficient algorithms to construct the semantic graph on a distributed computing environment (i.e. Spark) and conduct numerical data augmentation using the graph. Through comprehensive experiments on real-world datasets, the approach is shown to (1) achieve better accuracy on semantic matches and value conversions and (2) scales to the tractable computation time on large-scale data. Finally, we also present an interface to apply the semantic graph for real-world scenarios.