Mendeley Data Showcase

A dataset of 100 product design cases for evaluating Human-AI intent alignment in generative design. It contains fuzzy initial prompts, explicit hard constraints (product type, materials, style), and expert-annotated ground-truth design descriptions. Sourced from Behance and AI-synthesized designs, it serves as a testbed for LLM-based conceptual design, constraint satisfaction, and multi-turn interaction evaluation.

Herein, we report an optimized synthetic pathway and application of S-functionalized 3,7,10-trimethylsilatrane derivatives (Sil69, Sil266, and SilNXT) as silatrane coupling agents (SilCAs) dedicated to preparation of SSBR/BR/SiO2 composites. The tire performance of silatrane-based elastomers was compared to reference samples prepared with commercial triethoxysilyl compoundsSi 69, Si 266, and NXT. We discovered that the melting point of SilCAs can critically affect the degree of silica surface coverage during rubber mixing, as indicated by bound rubber content and the Payne effect. Furthermore, we proved vulcanization accelerating properties of the silatranyl moiety, as well as trialcoholamine through model reactions. The unique chemistry of the cage-shaped silica-binding group allowed us to achieve with SilNXT, 14% greater tensile strength, 8% higher wet traction, and comparable rolling resistance and abrasion resistance, in reference to NXT.

This dataset contains the data and source code supporting the results presented in the associated manuscript. The dataset includes BCLS data for different honeycomb topologies, training, validation, and test data for the MPNN surrogate model, BCLS prediction results, and the source code used for data preprocessing, model training, , and prediction.

Supplementary material for a clinical review article for the Journal of the American Academy of Dermatology (JAAD).

This dataset describes the diffraction efficiency, polarization angle and polarization ratio of the +1 order diffracted beam for ferroelectric and twisted nematic Liquid Crystal based Spatial Light Modulators (LCSLM) and for different orientations of the incident beam polarization. We consider SXGA-R3 Ferroelectric type LCSLM, and LC-R 720, and LC-R 1080 twisted nematic type LCSLM. The dataset also includes reflection efficiency, polarization angle and polarization ratio of the reflected beam for LC-R 720, and LC-R 1080 and for different orientations of the incident beam polarization. The measured data show that as the polarization of the incident beam is varied the diffraction efficiency in case of SXGA-R3 remains same although the polarization of the diffracted beam changes while in case of LC-R 720, and LC-R 1080 the behavior is opposite. The dataset (LCSLM-DiffractionEfficiency.zip) contains three csv files namely, SXGA-R3-DE.csv containing the data using the ferroelectric LCSLM (SXGA-R3), LC-R720-DE.csv containing the data using the twisted nematic LCSLM (LC-R 720) and LC-R1080-DE.csv containing the data using the nematic LCSLM (LC-R 1080). In each csv file column 1 displays the incident beam power, column 2 displays the polarization angle of the incident beam, column 3 displays the power in the +1 order beam without analyzer normalized by the incident power (i.e. diffraction efficiency), column 4 displays the dominant polarization direction in the +1 order beam, column 5 displays the normalized power in the +1 order beam with the analyzer in the dominant polarization direction, column 6 displays the normalized power in the +1 order beam with the analyzer perpendicular to the dominant polarization direction, and column 7 displays polarization ratio of the +1 order beam (i.e. column 5/ column 6). The dataset (LCSLM-ReflectionEfficiency.zip) has eight csv files namely, LC-R720-RE_g=i.csv where i=0, 50, 150, 255 corresponding to different grayscale values of the uniform pattern for the twisted nematic LCSLM (LC-R 720) and LC-R1080-RE_g=i.csv where i=0, 50, 150, 255 corresponding to different grayscale values of the uniform pattern for the twisted nematic LCSLM (LC-R 1080). In each csv file column 1 displays the incident beam power, column 2 displays the polarization angle of the incident beam, column 3 displays the power in the reflected beam without analyzer normalized by the incident power (i.e. reflection efficiency), column 4 displays the dominant polarization direction of the reflected beam, column 5 displays the normalized power of the reflected beam with the analyzer in the dominant polarization direction, column 6 displays the normalized power of the reflected beam with the analyzer perpendicular to the dominant polarization direction, and column 7 displays polarization ratio of the reflected beam (i.e. column 5/ column 6).

NeemLeaf is a dataset of field-acquired images of Azadirachta indica (Neem) designed for the assessment of the health of plants and for artificial intelligence. The dataset has a total of 4,550 images that are classified into the following six classes: Healthy Leaves (2,046), Diseased Leaves (1,010), Yellow Leaves (976), Young/Growing Leaves (400), Shadowed Leaves (94) and Neem Trees (24). The images were taken in the natural field environment of Bharuwadih Khurd Village, Tilda Development Block, Raipur District, Chhattisgarh, India using the Motorola Edge 50 Fusion smartphone. The dataset has variations in the field of different lighting conditions, backgrounds, appearances of leaves, and viewing conditions, and can support the research of computer vision, deep learning, the assessment of plant health, transfer learning, and precision agriculture.

Digital traceability can make environmental and social responsibility visible, yet restrained communication may limit how consumers use that evidence. Integrating signaling and attribution theories, this study examines how environmental and social digital traceability quality relates to corporate social responsibility authenticity, green trust, and green clothing purchase intention, with perceived greenhushing as a boundary condition. Survey data from 245 Vietnamese consumers were analyzed using partial least squares structural equation modeling and fuzzy-set qualitative comparative analysis. Environmental and social traceability were positively associated with authenticity, authenticity supported green trust, and green trust supported purchase intention. Perceived greenhushing weakened the trust relationship. Environmental traceability was core present in both high-intention solutions and core absent in both low-intention solutions. Together, the findings show that visibility, authenticity, and trust shape how traceability evidence supports sustainable purchase decisions.

This repository contains the article-focused supplementary software and audit materials supporting the manuscript “Lower Error Need Not Preserve Biological Fidelity: A Frozen Selective-Prediction Evaluation of Single-Cell Perturbations.” The package includes frozen protocols and configurations, scGen Article Gate I analysis code, validation tests, result receipts, figure-generation scripts and outputs, environment records, A/B-machine provenance, SHA256 manifests, reproduction instructions, and a BSD-3-Clause license. The frozen Article Gate I decision is NO-GO and remains unchanged. The repository does not include H5AD source data, Tahoe-100M, McFarland data, the separate SGM-DEGAWARE branch, OP3 full counts, or de_test.h5ad. Public source-data access is documented in the manuscript and package. The included materials support audit and reproducibility without authorising Gate II or additional scientific computation.

Dataset containing historical options data from FirstRateData's Historical Options Data Bundle (the downloadable sample).

The Smart Meter Data file contains all valid smart meter measurements for a settlement, with all anomalies and data gaps removed